US20050125307A1 - Approach for estimating user ratings of items - Google Patents
Approach for estimating user ratings of items Download PDFInfo
- Publication number
- US20050125307A1 US20050125307A1 US10/773,843 US77384304A US2005125307A1 US 20050125307 A1 US20050125307 A1 US 20050125307A1 US 77384304 A US77384304 A US 77384304A US 2005125307 A1 US2005125307 A1 US 2005125307A1
- Authority
- US
- United States
- Prior art keywords
- items
- users
- ratings
- recited
- particular item
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q30/00—Commerce
- G06Q30/06—Buying, selling or leasing transactions
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q30/00—Commerce
- G06Q30/02—Marketing; Price estimation or determination; Fundraising
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q30/00—Commerce
- G06Q30/02—Marketing; Price estimation or determination; Fundraising
- G06Q30/0201—Market modelling; Market analysis; Collecting market data
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q30/00—Commerce
- G06Q30/02—Marketing; Price estimation or determination; Fundraising
- G06Q30/0201—Market modelling; Market analysis; Collecting market data
- G06Q30/0204—Market segmentation
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q30/00—Commerce
- G06Q30/02—Marketing; Price estimation or determination; Fundraising
- G06Q30/0207—Discounts or incentives, e.g. coupons or rebates
- G06Q30/0217—Discounts or incentives, e.g. coupons or rebates involving input on products or services in exchange for incentives or rewards
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q30/00—Commerce
- G06Q30/06—Buying, selling or leasing transactions
- G06Q30/0601—Electronic shopping [e-shopping]
- G06Q30/0623—Item investigation
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q40/00—Finance; Insurance; Tax strategies; Processing of corporate or income taxes
- G06Q40/04—Trading; Exchange, e.g. stocks, commodities, derivatives or currency exchange
Definitions
- the present invention relates to inventory rental, and more specifically, to an approach for estimating user ratings of items.
- Conventional inventory rental models are typically based upon renting items for fixed rental periods and charging late fees for keeping rented items beyond a specified return date.
- the term “item” refers to any item of commerce.
- These types of inventory models suffer from several significant limitations.
- First, conventional rental models require customers to make the decision of what items to rent at substantially the same time as the decision of when to rent the items.
- An example that illustrates this limitation is a video rental business.
- Customers go to a video rental store and select particular movies to rent at that time. The customers take the movies home and must return them by a particular due date or be charged a late fee. In this situation, the customers cannot decide what movies to rent before actually renting them.
- the customers may have a particular movie in mind, but there is no guarantee that the video rental store has the particular movie in stock.
- due dates are inconvenient for customers, particularly for “new release” movies that are generally due back the next day.
- FIG. 1 is a diagram depicting an approach for renting items to customers according to an embodiment.
- FIG. 2 is a flow diagram depicting an approach for renting items to customers according to an embodiment.
- FIG. 3 is a flow diagram depicting a “Max Out” approach for renting items to customers according to an embodiment.
- FIG. 4 is a flow diagram depicting a “Max Turns” approach for renting items to customers according to an embodiment.
- FIG. 5 is a diagram depicting an approach for renting audio/video items to customers over the Internet according to an embodiment.
- FIG. 6 is a flow diagram illustrating an approach for renting audio/video items to customers over the Internet using both “Max Out” and “Max Turns” according to an embodiment.
- FIG. 7 is a flow diagram that depicts an approach for estimating how a user would rate an item according to an embodiment of the invention.
- FIG. 8 is a block diagram that depicts users' ratings of items arranged in an array.
- FIG. 9 is a block diagram that depicts an arrangement for recommending items according to an embodiment of the invention.
- FIG. 10 is a block diagram that depicts another arrangement for recommending items according to an embodiment of the invention.
- FIG. 11 is a block diagram of a computer system upon which embodiments of the invention may be implemented.
- FIG. 1 is a block diagram 100 that illustrates an approach for renting items to customers according to various embodiments described herein.
- the term “items” refers to any commercial goods that can be rented to customers. Examples of items include movies, music and games stored on a non-volatile memory such as a tape, other magnetic medium, optical medium, read-only memory or the like, and the invention is not limited to any particular type of item.
- the decision of what items to rent is separated from the decision of when to rent the items.
- Customers may specify what items to rent using one or more item selection criteria separate from deciding when to receive the specified items.
- customers are not constrained by conventional rental “windows” and instead can have continuous, serialized rental of items.
- a customer 102 provides one or more item selection criteria to a provider 104 over a link 106 .
- Link 106 may be any medium for transferring data between customer 102 and provider 104 and the invention is not limited to any particular medium. Examples of link 106 include, without limitation, a network such as a LAN, WAN or the Internet, a telecommunications link, a wire or optical link or a wireless connection.
- the item selection criteria indicate items that customer 102 desires to rent from provider 104 .
- provider 104 provides the items indicated by the item selection criteria to customer 102 over a delivery channel 108 .
- Delivery channel 108 may be implemented by any mechanism or medium that provides for the transfer of items from provider 104 to customer 102 and the invention is not limited to any particular type of delivery channel. Examples of delivery channel 108 include, without limitation, mail delivery, courier delivery or delivery using a delivery agent.
- Provider 104 may be centralized or distributed depending upon the requirements of a particular application.
- a “Max Out” approach allows up to a specified number of items to be rented simultaneously to customer 102 by provider 104 .
- a “Max Turns” approach allows up to a specified number of item exchanges to occur during a specified period of time.
- the “Max Out” and “Max Turns” approaches may be used together or separately with a variety of subscription methodologies.
- step 204 customer 102 creates item selection criteria.
- step 206 customer 102 provides the item selection criteria to provider 104 .
- step 208 in response to provider 104 receiving the item selection criteria from customer 102 , provider 104 provides one or more items indicated by the item selection criteria to customer 102 .
- the process is complete in step 210 .
- the one or more item selection criteria provided by customer 102 to provider 104 indicate the particular items that customer 102 desires to rent from provider 104 .
- the item selection criteria define a customer-specific order queue that is fulfilled by provider 104 .
- the item selection criteria specify attributes of items to be provided by provider 104 to customer 102 .
- Item selection criteria may specify any type of item attributes and the invention is not limited to particular item attributes. Examples of item attributes include, without limitation, identifier attributes, type attributes and cost attributes. Item selection criteria may be changed at any time to reflect changes in items that customers desire to rent from a provider.
- items are delivered by provider 104 to customer 102 over delivery channel 108 based upon item delivery criteria. More specifically, the delivery of items from provider 104 to customer 102 is triggered by item delivery criteria being satisfied.
- the item delivery criteria may include a wide range of criteria and the invention is not limited to any particular item delivery criteria. Examples of item delivery criteria include, without limitation, customer request/notification, customer notification that an item is being returned, customer return of an item, the occurrence of a specified date, the elapsing of a specified period of time or a customer payment.
- the item delivery criteria may be specified by customer 102 to provider 104 or negotiated by customer 102 and provider 104 as part of a subscription service.
- a particular subscription service may include item delivery criteria that specifies that a particular number of items are to be delivered monthly.
- item delivery criteria may specify that an initial set of items is to be delivered by provider 104 to customer 102 upon initiation of a subscription service and that additional items are to be delivered to customer 102 upon return of items to provider 104 .
- Item delivery criteria may be applied uniformly to all items to be delivered to a customer, or may be item specific.
- item delivery criteria may specify a particular date, i.e., the third Wednesday of every month, for all item deliveries. Alternatively, separate item delivery dates may be assigned to each item.
- a “Max Out” approach is used to manage the number of items that may be simultaneously rented to customers.
- up to a specified number of items may be rented simultaneously to a customer.
- the “Max Out” approach establishes the size of an inventory of items that may be maintained by customers.
- the specified number of items may be specific to each customer or may be common to one or more customers. In the present example, if the specified number of items is three, then up to three items may be rented simultaneously by provider 104 to customer 102 . If the specified number of items are currently rented to customer 102 and the specified item delivery criteria triggers the delivery of one or more additional items, then those items are not delivered until one or more items are returned by customer 102 to provider 104 .
- the specified number of items in situations where the specified number of items are currently rented to customer 102 and the specified item delivery criteria triggers the delivery of one or more additional items, then the one or more additional items are delivered to customer 102 and customer 102 and a surcharge is applied customer 102 .
- the specified number of items may then be increased thereafter to reflect the additional items delivered to customer 102 and increase the size of the inventory maintained by customer 102 .
- the specified number of items may remain the same and number of items maintained by customer 102 returned to the prior level after items are returned to provider 104 by customer 102 .
- the specified number of items may be unlimited.
- step 304 one or more initial items are delivered to customer 102 to establish the inventory maintained by customer 102 . Note that an initial delivery of items is not required and according to one embodiment, the inventory of customer 102 is incrementally established over time.
- step 306 a determination is made whether the item delivery criteria have been satisfied. If not, then the determination continues to be made until the item delivery criteria are satisfied.
- the delivery criteria may include customer notification generally, customer notification that an item is being returned, the actual return of an item, the occurrence of a specific date, or that a specified amount of time has elapsed.
- step 308 a determination is made whether the specified number of items have been delivered. If not, then control returns to step 304 and one or more additional items are delivered by provider 104 to customer 102 . If however, in step 308 , the specified number of items have been delivered, then in step 310 , a determination is made whether the specified number of items, i.e., the “Max Out” limit, is to be overridden. As previously described, the specified number of items may be overridden by increasing the specified number of items, i.e., the “Max Out” limit, to allow additional items to be delivered to customer 102 and charging a fee to customer 102 . Alternatively, the specified number of items is not changed and a surcharge applied to customer 102 . This process continues for the duration of the subscription and is then complete in step 312 .
- the specified number of items i.e., the “Max Out” limit
- a “Max Turns” approach is used to rent items to customers.
- up to a specified number of item exchanges may be performed during a specified period of time. For example, referring to FIG. 1 , suppose that provider 104 agrees to rent items to customer 102 with a “Max Turns” limit of three items per month. This means that customer 102 may make up to three item exchanges per month.
- This approach may be implemented independent of the number of items that a customer may have rented at any given time under the “Max Out” approach. The approach is also independent of the particular item delivery criteria used.
- the “Max Turns” approach is implemented in combination with the “Max Out” approach to rent items to customers. In this situation, up to a specified number of total items are simultaneously rented to customer 102 and up to a specified number of item exchanges may be made during a specified period of time.
- using the “Max Out” and the “Max Turns” approaches together essentially establishes a personal item inventory for customer 102 based upon the “Max Out” limit that may be periodically refreshed based upon the “Max Turns” limit selected.
- customer 102 may wish to exchange more than the specified number of items during a specified period.
- provider 104 agrees to rent additional items above the specified number to customer 102 and to charge customer 102 for the additional items. For example, suppose that provider 104 agrees to rent items to customer 102 with up to three item turns (exchanges) per month. If, in a particular month, customer 102 requires two additional turns, then the two additional items are provided to customer 102 and a surcharge is applied to customer 102 for the additional two items.
- customer 102 may not use all of its allotted turns during a specified period.
- customers lose unused turns during a subscription period. For example, if customer 102 has a “Max Turns” limit of four item exchanges per month and only makes two item exchanges in a particular month, then the two unused exchanges are lost and cannot be used. At the start of the next month, customer 102 would be entitled to four new item exchanges.
- customers are allowed to carry over unused turns to subsequent subscription periods. For example, if customer 102 has a “Max Turns” limit of four item exchanges per month and only makes two item exchanges in a particular month, then the two unused exchanges are lost and cannot be used. At the start of the next month, customer 102 would be entitled to six new item exchanges, two from the prior month and four for the current month.
- step 404 customer 102 and provider 104 agree upon the terms of the “Max Turns” agreement. Specifically, customer 102 and provider 104 agree at least upon the maximum number of turns that are allowed in a specified period of time.
- provider 104 in response to one or more item delivery criteria being satisfied, provides one or more items to customer 102 over delivery channel 108 .
- Any type of item delivery criteria may be used with the “Max Turns” approach and the invention is not limited to any particular delivery criteria.
- the initial one or more items may be delivered to customer 102 in response to a subscription payment made by customer 102 to provider 104 , the initiation of a specified subscription period, or by request of customer 102 for the initial-rental items.
- the number of initial one or more items must not exceed the terms of the “Max Turns”agreement.
- step 408 in response to one or more delivery criteria being satisfied, a determination is made whether additional items can be provided to customer 102 within the terms of the “Max Turns” agreement. For example, if the number of items rented to customer in the current subscription period is less than the agreed-upon “Max Turns,” then additional items can be rented to customer 102 within the terms of the “Max Turns” agreement. In this situation, this determination may be made in response to customer 102 returning one or more items to provider 104 , or by customer 102 requesting additional items.
- step 408 If, in step 408 , a determination is made that additional items can be rented to customer 102 within the terms of the “Max Turns” agreement, then control returns to step 406 where one or more additional items are rented to customer 102 . If however, in step 408 , a determination is made that additional items cannot be rented to customer 102 within the-terms of the “Max Turns” agreement, then in step 410 , a determination is made whether to override the current agreement terms. If so, then in step 412 , the agreement terms are changed to allow for a larger number of terms and customer 102 is charged accordingly, or the terms are left unchanged and a surcharge is applied for the additional items to be delivered. Control then returns to step 406 , where one or more additional items are delivered to customer 102 .
- step 410 If in step 410 , a determination is made that the current agreement is not to be overridden, then in step 414 , no items are delivered to customer 102 until the next subscription period. For example, the request for additional items may be received at the end of a subscription period and instead of renting the additional items immediately, they are instead delivered during the subsequent subscription period. Control then returns to step 406 where one or more additional items are rented to customer or the process is complete in step 416 .
- FIG. 5 is a diagram 500 that depicts a set of customers 502 that desire to rent A/V items from a provider 504 .
- Customers 502 communicate with provider 504 over links 506 , the global packet-switched network referred to as the “Internet,” and a link 510 .
- Links 506 and 510 may be any medium for transferring data between customers 502 and the Internet 508 and between the Internet 508 and provider 504 , respectively, and the invention is not limited to any particular medium.
- links 506 and 510 may be connections provided by one or more Internet Service Providers (ISPs) and customers 502 are configured with generic Internet web browsers.
- ISPs Internet Service Providers
- Links 506 and 510 may be secure or unsecured depending upon the requirements of a particular application.
- customers 502 enter into a rental agreement with provider 504 to rent audio/video (A/V) items 512 from provider 504 according to the “Max Out” and/or “Max Turns” approaches described herein.
- the invention is not limited to any particular approach for entering into the rental agreement.
- customers 502 and provider 504 may enter into a rental agreement by mail, telephone or over the Internet, by customers 502 logging into a web site associated with provider 504 .
- Customers 502 create and provide item selection criteria to provider 504 over links 506 and 510 and the Internet 508 .
- the invention is not limited to any particular approach for specifying and providing item selection criteria to provider 504 .
- customers 502 provide item selection criteria to provider 504 in one or more data files.
- customers 502 log onto a web site of provider 504 and use a graphical user interfaced (GUI) to specify attributes of the movies and music that customers desire to rent from provider 504 .
- GUI graphical user interfaced
- the item selection attributes may include any attributes that describe, at least in part, movies, games or music that customers 502 desire to rent.
- example attributes include, without limitation, title, category, director name, actor name and year of release.
- example attributes include, without limitation, title and category.
- example attributes include, without limitation, title, category, artist/group name and year of release.
- Customers 502 may identify specific movies or music by the item selection criteria, or may provide various attributes and allow provider 504 to automatically select particular movies and music that satisfy the attributes specified.
- customers 502 may specify item selection criteria that include horror movies released in 1999 and let provider 504 automatically select horror movies that were release in 1999.
- customers 502 may specify item selection criteria that include adventure movies starring Harrison Ford.
- Customers 502 may also specify an order or priority for the specified item selection criteria.
- customers 502 may specify specific movie titles and the order in which they want to receive them.
- customers 502 may specify that they want to receive a particular number of movies of different types.
- A/V items 512 are rented to customers 502 over delivery channels 514 in accordance with the terms of the rental agreement. Specifically, according to the “Max Out” approach described herein, an initial set of A/V items 512 , such as movies, games and music, are delivered to customers 502 over delivery channels 514 according to the terms of the rental agreement. Subsequent A/V items 512 are delivered whenever the specified item delivery criteria are satisfied.
- additional A/V items 512 may be delivered upon the return of one or more A/V items 512 to provider, a request from customers 502 , the arrival of a particular date, e.g., a specific day of the month, or the expiration of a specified period of time, e.g., fifteen days.
- the rental agreement between customers 502 and provider 504 may also specify a maximum number of turns in combination with the “Max Out” approach.
- a maximum number of turns restricts how quickly customers 502 may refresh their A/V item 512 inventories. For example, suppose that a particular customer 502 agrees with provider 504 to rent up to four movies with a maximum of four turns per month. Under this agreement, the particular customer 502 may maintain a personal inventory of up to four movies and rent four new movies per month. Thus, the particular customer 502 can completely “replace” his personal inventory once per month. If the particular customer 502 agreed to a maximum of up to eight turns per month, then the particular customer 502 would be able to completely replace his personal inventory twice per month.
- Provider 504 is illustrated as a single entity for purposes of explanation only. Provider 504 may be centralized or distributed depending upon the requirements of a particular application. For example, provider 504 may be a central warehouse from which all A/V items 512 are provided. Alternatively, provider 504 may be implemented by a network of distributed warehouses.
- FIG. 6 is a flow diagram that illustrates an approach for renting A/V items 512 , e.g., movies, to customers over a communications network such as the Internet using both “Max Out” and “Max Turns” according to an embodiment.
- a customer 502 enters into a rental agreement with provider 504 .
- customer 502 uses a generic web browser to access an Internet web site associated with provider 504 and enter into a rental agreement that specifies that customer 502 may maintain a personal inventory of four movies (“Max Out” of four) and receive up to four new movies per month (“Max Turns” of four).
- the rental agreement specifies that new movies will be delivered upon return of a rented movie from customer 502 , i.e., the delivery criteria is a return of a movie by the customer.
- customer 502 creates and provides movie selection criteria to provider 504 that indicates movies that customer 502 desires to rent.
- the movie selection criteria may specify particular movie titles that customer 502 desires to rent.
- the movie selection criteria may also specify an order or priority in which customer 502 wishes to rent the movies.
- the movie selection criteria may specify movie preferences for customer 502 , e.g., types of movies, directors, actors, or any other movie preferences or attributes.
- provider 504 automatically selects particular titles that satisfy the movie selection criteria.
- the movie selection criteria may specify a preference for action movies starring a particular actor, with a preference for “new release” movies.
- Provider 504 attempt to provide movies to customer 502 that best satisfy the preferences indicated by the movie selection criteria.
- one or more initial movies 512 are delivered to customer 502 over delivery channel 514 .
- the one or more initial movies 512 may be delivered to customer 502 via mail, courier, delivery agent or any other suitable means negotiated between customer 502 and provider and the invention is not limited to any particular type of delivery mechanism. For purposes of explanation only, is presumed in the present example that movies are mailed between customer 502 and provider 504 .
- the one or more initial movies 512 establish the personal movie inventory of customer 502 .
- Customer 502 may choose to receive any number of movies up to the “Max Out” limit of four movies. Typically, customer 502 will choose to initially receive four movies in the initial delivery.
- step 610 a determination is made whether any movies 512 have been returned by customer 502 to trigger another movie delivery.
- the delivery of additional movies is triggered by the receipt, e.g., via mail, of one or more movies from customer 502 .
- the delivery of additional movies 512 may also be triggered by a request from customer 502 for additional movies 512 .
- customer 502 may notify provider 504 via telephone, email or by accessing the web site associated with provider 504 .
- step 610 a determination is made that one or more movies 512 were received from customer 502 . If, in step 610 , a determination is made that one or more movies 512 were received from customer 502 , then in step 612 , a determination is made whether the maximum number of turns (“Max Turns”) limit has been reached for the current cycle. In the present example, a determination is made whether four or more movies have been mailed in the current month. If not, then control returns to step 608 , where one or more additional movies 512 are mailed to customer 502 via delivery channel 514 up to the “Max Out” limit of four.
- Max Turns the maximum number of turns
- step 612 a determination is made that the “Max Turns” limit has been met for the current cycle, i.e., in the present example, four movies 512 have been mailed to customer 502 in the current month
- step 614 a determination is made whether to override the current “Max Turns” limit. If so, then in step 616 , a surcharge is applied to customer 502 and control returns to step 608 where the additional movies 514 are mailed to customer 502 . If not, then in step 618 , a determination is made whether to continue the subscription service. If so, then no additional movies are mailed to customer 502 during the current cycle, e.g., the current month, and the control returns to step 610 . If, in step 618 , a determination is made that service is not to be continued, then the process is complete in step 620 .
- customer 502 may desire to increase or decrease the size of customer's 502 personal movie inventory by changing the current “Max Out” limit.
- customer 502 notifies provider 504 , e.g., by telephone, mail, email or by accessing the web site associated with provider 504 , that customer 502 wishes to change the “Max Out” limit.
- the movie rental agreement between customer 502 and provider 504 is then modified to reflect the change of the “Max Out” limit.
- additional movies 512 may be immediately mailed to customer 502 .
- the approach described herein for renting items to customers provides superior inventory management to prior approaches.
- the use of item selection criteria provides for efficient inventory management by allowing the greatest number of items to be rented at any given time.
- the greatest number of customers are provided with their most preferred items.
- customers may specify priorities for the items indicated by the item selection criteria. Thus, if a particular customer's first choice is not available, or already rented, then the item having the next highest priority can be rented to the particular customer.
- customers may indicate items that are not yet available for rent. Then, the items are delivered to customers when they become available.
- a particular customer 502 desires to rent an as-yet-unreleased movie entitled “ABC.”
- the particular customer 502 indicates this movie to provider 504 by the item selection criteria. Since the movie ABC is not yet available, it cannot be delivered to the particular customer 502 . However, when the movie ABC does become available, it can be shipped immediately to the particular customer 502 , as well as other customers 502 who may have also requested the movie. This allows provider 504 to maximize the number of items rented while ensuring that customers 502 are able to rent the highest priority items that they requested.
- as yet unknown items may also be rented by specifying attributes of the unknown items.
- the particular customer 502 may request to rent the next new movie of a particular director, for which the exact name is unknown to the particular customer.
- the particular customer 502 may request to rent the next album of a particular group that is currently in process and does not yet have a title.
- a user would rate an item In some situations, it is desirable to know how a user would rate an item. For example, in the context of recommending items to users, it is useful to know how a user would rate an item, so that a decision can be made to either recommend or not recommend the item to the user. A recommendation is more valuable to the user if it is likely that the user would rate the item highly because the user is more likely to be happy with the recommendation.
- An approach is provided for estimating how a user would rate an item. The approach generally involves estimating how a user would rate an item based upon similarities between items and similarities of ratings given to items by both the user and one or more other users. The approach is applicable to any type of item, for example rental items such as movies and games, and the invention is not limited to any particular type of item. The approach is applicable to a wide variety of contexts and may be used, for example in a recommendation engine to recommend items to users.
- FIG. 7 is a flow diagram 700 that depicts an approach for estimating how a user would rate a particular item according to an embodiment of the invention.
- step 702 one or more items from a plurality of items are identified that have ratings similar to the particular item.
- step 704 one or more other users are identified that have given ratings to the one or more other items that are substantially similar to ratings given by the user to the one or more other items, and have also rated the particular item.
- an estimation of how the user would rate the particular item is generated based upon ratings given by the one or more other users for the particular item.
- FIG. 8 is a block diagram that depicts users' ratings of items arranged in an array 800 .
- Array 800 includes columns of items 802 and rows of users 804 .
- Each cell may be empty or contain data indicating that the user has not provided a rating for the item.
- a specified value such as a null value or other value, may be used to indicate that a user has not provided a rating for an item.
- each cell may contain data that represents a user rating of an item.
- Different types of data and values may be used to indicate a user rating of an item, depending upon the requirements of a particular application, and the invention is not limited to any particular data types or values.
- ratings may be represented by numerical data, e.g., in a range of 1 to 5 that corresponds to a 5-star rating scale or binary data that corresponds to a favorable or unfavorable rating.
- Embodiments of the invention are described herein in the context of users' ratings in array 800 for explanation purposes only and the invention is not limited to users' ratings being maintained in any particular format or data structure.
- the relative size of the items 802 and users 804 dimensions depicted in FIG. 8 is not meant to imply that there must be more users 804 than items 802 .
- Embodiments of the invention are described herein in the context of generating an estimated rating for a particular item 806 , from items 802 , and a particular user 808 , from users 804 .
- the estimated rating for the particular item 806 and particular user 808 is indicated in FIG. 8 by reference numeral 810 .
- a set of similar items 812 that have ratings similar to the particular item 806 is identified. Similar items 812 are depicted in FIG. 8 as being located adjacent to the particular item 806 in array 800 for explanation purposes only. Similar items 812 may include any of items 802 . A variety of techniques may be used to determine similar items 812 , depending upon the requirements of a particular application, and the invention is not limited to any particular technique.
- similar items 812 are items from items 802 that satisfy item similarity criteria with respect to particular item 806 .
- other items that satisfy the item similarity criteria with respect to the particular item are considered to be similar to the particular item.
- Similarity is not necessarily symmetrical.
- an item A may be considered similar to particular item 806 and therefore included in similar items 812 for particular item 806
- particular item 806 may, or may not be, sufficiently similar to item A to be considered similar to item A and included in a set of similar items for item A.
- the number of items in similar items 812 may be limited to a specified number to reduce the amount of computational resources and time required to generate an estimated rating for particular item 806 .
- the specified number of items in similar items 812 may represent the first specified number of items that satisfied the item similarity criteria for particular item 806 , or may represent the specified number of items that satisfied the item similarity criteria for particular item 806 with similarity above some specified level or threshold.
- the item similarity criteria may include a wide variety of criteria, depending upon the requirements of a particular implementation, and the invention is not limited to any particular item similarity criteria.
- One example of item similarity criteria is a statistical correlation threshold. Thus, two items that satisfy the statistical correlation threshold are considered to be similar to each other.
- a wide variety of statistical measures may be used for this purpose, depending upon the requirements of a particular implementation, and the invention is not limited to any particular statistical measure of similarity.
- X i are the ratings for one item or user
- Y i are the ratings for the other item or user
- Table I contains example user ratings for items X and Y, the product of the user ratings for X and Y, the square of the user ratings for X and Y and the summations of these values:+ TABLE I X Y X ⁇ Y X**2 Y**2 1 2 2 1 4 4 4 1 4 16 1 5 4 20 25 16 3 3 9 9 9 3 5 4 20 25 16 1 4 4 16 16 16 4 5 20 16 25 Summations 26 23 91 108 87
- the item similarity criteria is a specified number of co-ratings between items.
- a co-rating exists for two items when a user rates both items.
- the item similarity criteria may specify a minimum number of co-ratings that must be shared between two items for the two items to be considered similar items.
- Yet another example of the item similarity criteria is a specified fraction of co-ratings between items divided by the number of ratings of either item.
- the item similarity criteria may specify a minimum fraction of co-ratings between two items divided by the number of ratings of either item that must be satisfied for the two items to be considered similar items.
- E ⁇ ( R ab ) U ⁇ ⁇ • ⁇ ⁇ R a U ⁇ ⁇ • ⁇ ⁇ R b U If the actual number of co-raters of both Item A and Item B R ab is much larger than E(R ab ), it is likely that Item A and Item B are related in some way.
- R ab is much smaller than E(R ab )
- Item A and Item B are dissimilar in a way such that raters of Item A are unlikely to encounter Item B, or unlikely to have an opinion about Item B and so are less likely to rate Item B (and vice-versa). Either of these pieces of information may be used as additional similarity criteria. Note: if one of Item A or Item B is newer than the other (in other words, did not exist at the time the other Item was popular or well-known), then it is very likely that the total number of co-ratings of the Items will be much smaller than would be expected by the basic formula above.
- the pool of ratings may be subdivided into slices by time—for example, months—and the co-raters fraction calculation described above may be applied to in each successive (month) slice.
- the ratings distributions of the two items are “factored out.” This is performed by taking the percentage of each rating for each item, and multiplying it out in a two-dimensional array to obtain expected percentages for every possible pair of ratings (r 1 , r 2 ) for the two items. Multiplying each of the array values by the total number of co-raters produces the expected count of co-raters for each rating pair, if the items are unrelated. The actual counts of co-ratings in each of the array cells are compared to see if they are more or less than expected. Higher than expected values in areas corresponding to similarity, e.g. a user that rated both items high or both item slow, indicates that the items are similar. Higher than expected co-rater counts in areas where a user liked one item and disliked another indicates that the items are not positively related. This use of the underlying ratings distribution for all raters provides a better understanding of values for co-raters.
- the Pearson Correlation Coefficient is required to be greater than a threshold (e.g. 0.3), and also the fraction of co-raters R ab must be at least some fraction (e.g. one half) the expected number of co-raters E(Rab), and the absolute number of co-raters must be at least some minimum number (e.g. 200).
- the methods described above show how to calculate the similarity between two items in order to find a set of similar items.
- the exact same methods may be applied in order to calculate the similarity between two users in order to find a set of similar users.
- items are additionally classified by semantic feature values.
- the semantic features might be genre (or category), director, lead actors/actresses, other actors/actresses, etc.
- the feature values are the genre, director, or actor names.
- the features might be styles, soloists, musicians, and lyricists.
- Each type of feature is given an importance score, which may be positive, negative, or zero, and two items may be compared by counting the number of feature matches and summing their importance scores.
- the absence of matches may be accorded a negative score (for example, the absence of matching directors might score the pairing ⁇ 0.1 points).
- the set of similar items is the specified number of items whose similarity score with the particular item is largest, or the set of items whose similarity score is above a specified threshold.
- tools may be provided for domain experts to specify item similarity by non-machine-implemented means. Once additional ratings data becomes available, the similar items specified by domain experts may be supplemented or discarded.
- one or more users from users 804 that have given ratings to similar items 812 that are similar to ratings given by the particular user 808 to similar items 812 are identified. All of users 804 may be evaluated for this purpose. There may be a significant and unacceptable computational cost, however, in evaluating all of users 804 in some applications where the number of users 804 is large. Furthermore, some users may be less useful than others because they have very little ratings data or have ratings data with poor statistical characteristics, as described in more detail hereinafter.
- a subset of users 804 is determined based upon specified user selection criteria.
- Reference users 814 are users from all users 804 that satisfy the specified user selection criteria.
- a wide variety of user selection criteria may be used, depending upon the requirements of a particular application, and the invention is not limited to any particular user selection criteria.
- the user selection criteria include a minimum number of ratings. Thus, users from users 804 that do not have at least the minimum number of ratings are not included in reference users 814 .
- the user selection criteria may include a maximum number of ratings. Thus, users from users 804 that have more than the maximum number of ratings are not included in reference users 814 .
- the user selection criteria specify an average rating range.
- users from users 804 with an average rating that is outside the specified range are not included in reference users 814 .
- ratings are provided on a numerical scale from one to five.
- the user selection criteria specify that the acceptable average rating range is two to four. In this situation, users 804 with an average rating that is less than two or greater than four are not included in reference users 814 .
- the user selection criteria specify that user ratings must conform to within a specified tolerance of a Normal distribution.
- users from users 804 with a rating distribution that does not approximate a Normal distribution within a specified tolerance are not included in reference users 814 .
- ratings are on a whole number numerical scale of one to five.
- a user that has more one ratings than two ratings, more two ratings than three ratings, more five ratings than four ratings, or more four ratings than three ratings is not included in reference users 814 .
- the specified tolerance may vary, depending upon the requirements of a particular application. For example, some applications may use a specified tolerance that allows a user in reference users 814 to have a slightly greater number of two ratings than three ratings, and an equal number of two ratings and one ratings, but not more one ratings than three ratings.
- the user selection criteria specify a range of acceptable standard deviations of user ratings.
- users from users 804 with a standard deviation of ratings that is not within the range of acceptable range of standard deviations of ratings are not included in reference users 814 .
- the user selection criteria might specify a standard deviation range of between 10% and 20%.
- a user having a standard deviation of ratings of less than 10% or 20% or greater is not included in reference users 814 .
- Generating an estimation of how particular user 808 would rate particular item 806 generally involves identifying which users 804 have rated similar items 812 items similarly to particular user 808 , and then generating the estimation based upon how those users 804 rated the particular item 806 .
- the approach generally involves identifying which reference users 814 have rated similar items 812 items similarly to particular user 808 , and then generating the estimation based upon how those reference users 814 rated the particular item 806 .
- Embodiments of generating an estimation of how particular user 808 would rate particular item 806 are described herein in the context of using reference users 814 , although the invention is not limited to this context.
- reference users' 814 ratings of similar items 812 are compared to the particular user's 808 ratings of similar items 812 to identify reference users 814 that are the most similar to particular user 808 .
- reference users 814 that have provided ratings for similar items 812 that are most similar to the ratings for similar items 812 provided by particular user 808 are referred to hereinafter as similar reference users 816 .
- the horizontal lines depicted across similar items 812 for the similar reference users 816 are provided for illustration purposes only and are not meant to imply that the similar reference users 816 have necessarily provided ratings for all similar items 812 .
- rating similarity criteria are used to identify the similar reference users 816 .
- the rating similarity criteria may be similar to the item similarity criteria described herein.
- the similar reference users 816 may be determined based upon determining statistical correlations between the ratings of similar items 812 provided by reference users 814 and the particular user 808 and then selecting the similar reference users 816 based upon the reference users 814 that best correlate to the particular user 808 .
- the similarity between similar items 812 and the particular item 806 may also be considered in identifying the similar reference users 816 . More specifically, ratings of similar items 812 that are more similar to the particular item 806 may be given more weight in determining the similar reference users 816 from reference users 814 .
- the estimation of how the particular user 808 would rate the particular item 806 is generated based upon the ratings of the particular item 806 made by the similar reference users 816 .
- the degree of similarity between the similar reference users 816 and the particular user 808 may vary based upon a variety of factors including, for example, the number of similar items 812 that both the particular user 808 and the similar reference users 816 have rated and how the particular user 808 and the similar reference users 816 rated the similar items 812 .
- the estimation may be generated in different ways, depending upon the similarity between the ratings of similar items 812 provided by the similar reference users 816 and the particular user 808 .
- the ratings provided by the one or more similar reference users 816 for similar items 812 are identical to the ratings provided by the particular user 808 for the same similar items 812 . This may occur, for example, where the number of similar items 812 rated by both the similar reference users 816 and the particular user 808 is relatively small.
- the estimated rating of the particular item 806 by the particular user 808 may be the average rating made by the one or more similar reference users 816 for the particular item 806 .
- the estimated rating of the particular item 806 by the particular user 808 may be the weighted linear sum of the ratings made by the one or more similar reference users 816 for the particular item 806 .
- the weighting applied varies based upon the similarity between each similar reference users 816 and the particular user 808 .
- the weighted linear sum of the ratings for particular item 806 is the sum of the products of the weights and the similar reference users' 816 ratings of particular item 806 , divided by the sum of the weights.
- the ratings of particular item 806 provided by the similar reference users 816 that are most similar to the particular user 808 are given more weight than the ratings of particular item 806 provided by the similar reference users 816 that are less similar to the particular user 808 .
- the similar reference users 816 include users A and B that are determined to have similarities to the particular user 808 of 0.8 and 0.5 (on a scale of 0 to 1), respectively.
- users A and B have given ratings for particular item 806 of 5 and 3 (on a scale of 1 to 5), respectively.
- the average rating of particular item 806 by users A and B is 4 ((5+3)/2).
- the weighted linear sum of the ratings for particular item 806 by users A and B is 4.23 ((5*0.8+3*0.5)/1.3).
- the weighted linear sum of the ratings is greater than the average of the ratings in this situation because the higher rating of user A is given greater weight than the lower weighting of user B, by virtue of the relatively greater similarity between user A and the particular user 808 than between user B and the particular user 808 .
- Various confidence metrics may be calculated that indicate a confidence level in an estimated rating of particular item 806 by the particular user 808 .
- a confidence metric is calculated based upon the disparity or spread between estimated ratings, where each estimated rating is based upon a different one of the similar reference users 816 . The standard deviation of the estimated ratings may be used for this purpose.
- the calculation of confidence based upon the disparity between estimated ratings may weight these estimated ratings based on the strength of the similarity between the reference user 816 and the particular user 808 .
- a confidence metric may also be made based upon similarities between the particular user 808 and each similar reference user 816 that is used to estimate a rating. The calculation of confidence may also take into account the absolute strength of the similarity between the reference users 816 used to generate the prediction and the particular user 808 .
- a weighted average of all reference users 814 who each share at least a rating of one of similar items 812 with the particular user 808 and who have each rated the particular item 806 , may be used to generate an estimation of how the particular user 808 would rate the particular item 806 .
- the weightings are applied based upon the similarity of each reference user 814 to the particular user 808 .
- Confidence metrics may also be applied to generating an estimation of how the particular user 808 would rate the particular item 806 based upon all reference users 814 used to generate the estimate.
- an average rating of the particular item 806 by all reference users 814 who have rated the particular item 806 may be used as an estimate of how the particular user 808 would rate the particular item 806 .
- an average rating of all users 804 who have rated the particular item 806 may be used as an estimate of how the particular user 808 would rate the particular item 806 .
- a weighted average of all reference users 814 who each share at least a rating of one of similar items 812 with the particular user 808 and who have each rated the particular item 806 may be used.
- FIG. 9 is a block diagram that depicts an arrangement 900 for recommending items according to an embodiment of the invention.
- Arrangement 900 includes a Web server 902 communicatively coupled to a recommendation system 904 .
- recommendation system 904 is configured to process ratings data using prioritized tasks to provide improved availability and accuracy of item recommendations.
- Recommendation system 904 includes a recommendation engine 906 , a non-volatile storage 908 and a ratings database (DB) 910 .
- DB ratings database
- Recommendation engine 906 may be implemented by any mechanism or any combination of hardware and software for performing the functions described herein. For purposes of explanation only, recommendation engine 906 is described in the context of a software process executing in a volatile memory.
- Non-volatile storage 908 may be implemented by any type of non-volatile storage mechanism and the invention is not limited to any particular mechanism.
- Example implementations include, without limitation, non-volatile memory and disc storage.
- Ratings DB 910 may be implemented by any database management mechanism and the invention is not limited to any particular implementation.
- Ratings and prediction data 912 may be maintained in a volatile memory, such as a random access memory (RAM), to reduce access times.
- ratings and prediction data 912 is maintained in a RAM array and is addressed using 64 bit addressing as provided by high-end computing architectures. The array is laid out such that two bytes are used for each cell, containing a rating value, an estimated rating value, a confidence value, and one or more flags indicating the presence and validity of the values.
- the RAM array may be considered as divided into a number of pages of perhaps 8,192 bytes each (which corresponds to the physical structure of current computers). The cells are laid out so that each user's ratings and estimated ratings span one or more pages for locality of access.
- ratings and prediction data 912 is periodically saved to a data file 914 maintained on non-volatile storage 908 .
- low-level operating system primitives may be used to ensure that data written to the RAM array is also stored in data file 914 in a timely manner, so that recommendation engine 906 may be periodically shut down for maintenance and upgrade without losing any of the ratings and prediction data 912 .
- data from data file 914 is mapped back (reloaded) into the RAM array, thereby restoring all the results of previous computations.
- the memory mapping functionality of the Solaris operating system offered by Sun Microsystems, Inc. permits data file 914 to be sparsely populated with disc blocks, such that any pages in the RAM array that have no values do not occupy space on non-volatile storage 908 .
- This provides considerable savings in hardware cost, since much of the RAM array will never contain ratings and prediction data.
- Ratings data is also stored in ratings DB 910 , which allows data file 914 to be re-generated in the event of a failure of non-volatile storage. Ratings DB 910 may also store any data contained in the RAM array or data file 914 or any other data used by recommendation system 904 , depending upon the requirements of a particular implementation.
- data file 914 also allows storage of more ratings and prediction data 912 than can be stored in the RAM array, given practical limitations in available RAM on computer architectures.
- the size of the RAM array is selected to be large enough to store ratings and prediction data 912 for at least reference users 814 and active users of recommendation system 904 . As users' status changes from active to inactive, the physical pages of storage in the RAM array allocated to their rating and prediction data 912 may be reclaimed and reused for other users just becoming active.
- Recommendation engine 906 is configured to perform a wide variety of tasks. These tasks may be prioritized in one or more task queues 916 to coordinate the activity of one or more processing threads.
- the one or more task queues may be stored in a volatile storage, such as RAM, or in non-volatile storage 908 .
- the processing threads process tasks from the task queues 916 based upon priority and the processing of a task may create additional queue requests that are added to task queues 916 .
- recommendation engine 906 is configured to process requests for estimated ratings and requests for recommendations.
- An estimated rating for a particular item for a particular user may be requested in situations where, for example, a user is viewing data associated with the particular item, such as a review, on a graphical user interface (GUI), and it is desirable to display data that indicates an estimate of how the particular user would rate the particular item.
- a request for recommendations may be made, for example, in situations where a user has requested a list of recommended items from a Web application.
- recommendation engine determines whether the requested estimated rating has previously been calculated and stored in the RAM array. If so, then the estimated rating is immediately returned to Web server 902 .
- a request is queued on the task queue with high priority to calculate the estimated rating (according to the algorithms described herein, or otherwise). If the compute threads complete the calculation of the estimated rating within some small threshold of time (for example, half a second) then the calculated estimated rating is returned. Otherwise, the recommendation engine 906 is configured to provide an average rating of the particular item as the estimated rating.
- recommendation engine 906 identifies the set of items with the highest estimated ratings and highest confidence for that user and returns to Web server 902 data that identifies the set of items as recommendations. If there are no confident predictions, then no recommendations may be returned.
- Recommendation system 904 may be used with any number of Web servers and also other types of computing entities.
- Recommendation engine 906 is also configured to receive and process rating data from Web server 902 .
- rating data which may include a new or changed rating
- recommendation engine 906 stores the rating in the RAM array using a simple address calculation and a simple memory write.
- the rating is also queued in a task queue 916 to be written to ratings DB 910 .
- estimated ratings for items similar to the item associated with the received rating data may need to be updated to reflect the received rating data. Therefore, one or more tasks are added to task queues 916 to recalculate the estimated ratings for items for which the item associated with the received rating data is considered similar.
- these tasks are given high priority, since the updated estimated ratings may be required in a relatively short time, for example for a merchandising decision or a recommendation for a user who is presumably visiting the Website in order to provide the rating.
- Quality metrics may be calculated from comparisons of actual user ratings to estimated user ratings. For example, suppose that the estimated rating 810 has been determined for the particular item 806 and the particular user 808 . Suppose further that this estimated rating 810 has been provided to Web server 902 , displayed for the particular user 808 , and the particular user 808 has provided an actual rating for the particular item 806 . A quality metric may be calculated based upon a comparison of the actual rating to the estimated rating 810 . This quality metric, along with other quality metrics, may be used to fine tune future estimated ratings.
- Recommendation engine 906 is further configured to periodically, and continuously if necessary, recalculate ratings and prediction data 912 .
- recommendation engine 906 is configured to recalculate the set of similar items for each item to reflect changes in ratings data and new ratings data. Changes in ratings data and new ratings data may be tracked and various thresholds and aging schemes used to prioritize recalculation of similar item sets. For example, items that have a relatively large amount of ratings data updated may be given priority over an item for which the calculated similar item set is simply old. It is beneficial to periodically recalculate the set of similar items for a particular item, even if no ratings data has changed for the particular item, since ratings data may have changed the set of items considered similar to the particular item.
- recommendation system 904 There may be situations where limitations in the power of the computing architecture used to implement recommendation system 904 make it difficult for recommendation system 904 to keep up with changes in ratings data. That is, the estimated ratings maintained by recommendation engine 906 may fall behind the updated ratings data. One consequence of this situation may be a loss in accuracy of estimated ratings. There may also be situations where limitations in the available size of non-volatile storage 908 may make maintaining data file 914 difficult. Accordingly, according to another embodiment of the invention, the approach is segmented into a series of separate recommendation engines and separate disc arrays allowing scaling to much higher capacities and load levels. FIG. 10 depicts an example embodiment of such an arrangement 1000 .
- Arrangement 1000 includes a Web server 1002 and a recommendation system 1004 , configured in accordance with an embodiment of the invention.
- Recommendation system 1004 includes a similars/reference engine 1006 , a non-volatile storage 1008 , a ratings DB 1010 and recommendation engines 1012 , 1014 .
- a single reference users/similar items data file 1016 is used to store the ratings for reference users 814 .
- Reference users/similar items data file 1016 is maintained on non-volatile storage 1008 by similars/reference engine 1006 , which maintains the set of reference users 814 .
- Reference users 814 are chosen as described herein. If too many of users 804 qualify as reference users 814 , then users 804 who have the most attractive rating patterns are chosen as reference users 814 .
- Example criteria for choosing the reference users 814 include, without limitation, users with the most ratings, or the most ratings on otherwise unrated items, or other criteria that may be appropriate.
- New candidate reference users are loaded from ratings DB 1010 as they qualify, and existing reference users 814 may be displaced if any of the candidate reference users have more attractive parameters making them more suitable as references users 814 for making estimated ratings.
- Similars/reference engine 1006 is also configured to calculate sets of similar items 812 for each of items 802 on a continuous basis and store that data into reference users/similar items data file 1016 .
- Recommendation engines 1012 , 1014 are configured to generate estimates of how users would rate items, as previously described herein. Recommendation engines 1012 , 1014 have exclusive read/write access to active user data files 1018 and 1020 , respectively. Active user data files 1018 , 1020 store ratings and prediction data for the users assigned to recommendation engines 1012 , 1014 . Recommendation engines 1012 , 1014 also have read-only access to reference users/similar items data file 1016 . Read-only access is inexpensive in resources on the disc system and computer IO systems except when the recommendation engines 1012 , 1014 are being started up.
- the reference users/similar items data file 1016 is periodically copied one or more times, and the resulting copy or copies are mapped by the recommendation engines 1012 , 1024 .
- Recommendation engines 1012 , 1014 estimate how users would rate items as described herein, obtaining the reference rating values from reference users/similar items data file 1016 and storing estimated ratings in their private read-write active user data files 1018 , 1020 .
- the assignment of users to recommendation engines 1012 , 1014 may be made using a variety of techniques, depending upon the requirements of a particular implementation, and the invention is not limited to any particular approach or assignment of users.
- the assignment of users to recommendation engines 1012 , 1014 may be dynamic and change over time, depending, for example, on loading requirements.
- Reference Users' 814 active profiles of ratings and predictions are stored in one of the active user data files 1018 , 1020 and estimated ratings for reference users 814 are calculated by the appropriate recommendation engine 1012 , 1014 and stored in the corresponding active user data files 1018 , 1020 .
- those ratings are also stored in ratings DB 1010 .
- similars/reference engine 1006 updates the shared copy of the ratings for that reference user 814 in reference user/similar items data file 1016 .
- Recommendation engines 1012 , 1014 may store ratings and prediction data in one or more local RAM arrays and then periodically write the ratings and prediction data to active user data files 1018 , 1020 .
- Task queues may be shared between the one or more engines, so that new work items may be triggered by one recommendation engine, but performed by a different engine if the task relates to data managed by the other engine.
- each portion of the ratings data for each reference user 814 is available to each similars/reference engine. More specifically, each data file is owned by a single similars/reference engine that has read/write access for maintenance, and each data file is also made available on a read-only basis to all other similars/reference engine for access to the reference ratings data.
- the approach described herein for estimating how a user would rate a particular item is applicable to any type of items or content and is well suited for Internet-based rental applications for renting movies, music and games to customers.
- the invention may be implemented in hardware circuitry, in computer software, or a combination of hardware circuitry and computer software and is not limited to a particular hardware or software implementation.
- the invention is ideally suited for recommendation engine applications and may be implemented as a stand-alone mechanism, or integrated into a rental mechanism.
- the approach is ideally suited for identifying rental items, such as movies, music and games, to recommend to users.
- FIG. 11 is a block diagram that illustrates a computer system 1100 upon which an embodiment of the invention may be implemented.
- Computer system 1100 includes a bus 1102 or other communication mechanism for communicating information, and a processor 1104 coupled with bus 1102 for processing information.
- Computer system 1100 also includes a main memory 1106 , such as a random access memory (RAM) or other dynamic storage device, coupled to bus 1102 for storing information and instructions to be executed by processor 1104 .
- Main memory 1106 also may be used for storing temporary variables or other intermediate information during execution of instructions to be executed by processor 1104 .
- Computer system 1100 further includes a read only memory (ROM) 1108 or other static storage device coupled to bus 1102 for storing static information and instructions for processor 1104 .
- ROM read only memory
- a storage device 1110 such as a magnetic disk or optical disk, is provided and coupled to bus 1102 for storing information and instructions.
- Computer system 1100 may be coupled via bus 1102 to a display 1112 , such as a cathode ray tube (CRT), for displaying information to a computer user.
- a display 1112 such as a cathode ray tube (CRT)
- An input device 1114 is coupled to bus 1102 for communicating information and command selections to processor 1104 .
- cursor control 1116 is Another type of user input device
- cursor control 1116 such as a mouse, a trackball, or cursor direction keys for communicating direction information and command selections to processor 1104 and for controlling cursor movement on display 1112 .
- This input device typically has two degrees of freedom in two axes, a first axis (e.g., x) and a second axis (e.g., y), that allows the device to specify positions in a plane.
- the invention is related to the use of computer system 1100 for implementing the techniques described herein. According to one embodiment of the invention, those techniques are performed by computer system 1100 in response to processor 1104 executing one or more sequences of one or more instructions contained in main memory 1106 . Such instructions may be read into main memory 1106 from another machine-readable medium, such as storage device 1110 . Execution of the sequences of instructions contained in main memory 1106 causes processor 1104 to perform the process steps described herein. In alternative embodiments, hard-wired circuitry may be used in place of or in combination with software instructions to implement the invention. Thus, embodiments of the invention are not limited to any specific combination of hardware circuitry and software.
- machine-readable medium refers to any medium that participates in providing data that causes a machine to operation in a specific fashion.
- various machine-readable media are involved, for example, in providing instructions to processor 1104 for execution.
- Such a medium may take many forms, including but not limited to, non-volatile media, volatile media, and transmission media.
- Non-volatile media includes, for example, optical or magnetic disks, such as storage device 1110 .
- Volatile media includes dynamic memory, such as main memory 1106 .
- Transmission media includes coaxial cables, copper wire and fiber optics, including the wires that comprise bus 1102 . Transmission media can also take the form of acoustic or light waves, such as those generated during radio wave and infrared data communications.
- Machine-readable media include, for example, a floppy disk, a flexible disk, hard disk, magnetic tape, or any other magnetic medium, a CD-ROM, any other optical medium, punchcards, papertape, any other physical medium with patterns of holes, a RAM, a PROM, and EPROM, a FLASH-EPROM, any other memory chip or cartridge, a carrier wave as described hereinafter, or any other medium from which a computer can read.
- Various forms of machine-readable media may be involved in carrying one or more sequences of one or more instructions to processor 1104 for execution.
- the instructions may initially be carried on a magnetic disk of a remote computer.
- the remote computer can load the instructions into its dynamic memory and send the instructions over a telephone line using a modem.
- a modem local to computer system 1100 can receive the data on the telephone line and use an infrared transmitter to convert the data to an infrared signal.
- An infrared detector can receive the data carried in the infrared signal and appropriate circuitry can place the data on bus 1102 .
- Bus 1102 carries the data to main memory 1106 , from which processor 1104 retrieves and executes the instructions.
- the instructions received by main memory 1106 may optionally be stored on storage device 1110 either before or after execution by processor 1104 .
- Computer system 1100 also includes a communication interface 1118 coupled to bus 1102 .
- Communication interface 1118 provides a two-way data communication coupling to a network link 1120 that is connected to a local network 1122 .
- communication interface 1118 may be an integrated services digital network (ISDN) card or a modem to provide a data communication connection to a corresponding type of telephone line.
- ISDN integrated services digital network
- communication interface 1118 may be a local area network (LAN) card to provide a data communication connection to a compatible LAN.
- LAN local area network
- Wireless links may also be implemented.
- communication interface 1118 sends and receives electrical, electromagnetic or optical signals that carry digital data streams representing various types of information.
- Network link 1120 typically provides data communication through one or more networks to other data devices.
- network link 1120 may provide a connection through local network 1122 to a host computer 1124 or to data equipment operated by an Internet Service Provider (ISP) 1126 .
- ISP 1126 in turn provides data communication services through the worldwide packet data communication network now commonly referred to as the “Internet” 1128 .
- Internet 1128 uses electrical, electromagnetic or optical signals that carry digital data streams.
- the signals through the various networks and the signals on network link 1120 and through communication interface 1118 which carry the digital data to and from computer system 1100 , are exemplary forms of carrier waves transporting the information.
- Computer system 1100 can send messages and receive data, including program code, through the network(s), network link 1120 and communication interface 1118 .
- a server 1130 might transmit a requested code for an application program through Internet 1128 , ISP 1126 , local network 1122 and communication interface 1118 .
- the received code may be executed by processor 1104 as it is received, and/or stored in storage device 1110 , or other non-volatile storage for later execution. In this manner, computer system 1100 may obtain application code in the form of a carrier wave.
- the novel approach described herein for renting items to customers allows the decision of what items to rent to be separated from the decision of when to rent the items.
- Customers may specify what items to rent using the item selection criteria and receive the items at a future point in time, without having to go to the provider to pick up the items.
- the selection criteria may be user specific and may indicate a desired fulfillment sequence.
- customers are not constrained by conventional due dates and instead may establish continuous, serialized rental streams of items.
- the approach also allows more efficient inventory management.
- the “Max Out” approach for inventory management allows users to maintain their own inventory of items that are periodically replaced by other items according to specified event criteria.
- the event criteria that trigger sending another item to a customer are very flexible and may be tailored to the requirements of a particular application.
- the event criteria may include a return of any of the items currently in use by the customer or merely customer notification. This is very convenient in the context of movie rentals since the return of a movie to the provider automatically triggers the sending of another movie to the customer.
- the “Max Turns” approach for inventory management, when used alone or in combination with “Max Out,” provides even greater flexibility for customers and providers. The max number of turns can be selected individually for particular customers depending upon their particular needs.
- the “Max Out” and “Max Turns” approaches provide great flexibility in establishing subscription models to satisfy the needs of a particular application. Specifically, the size and replacement frequency of customer inventories can be tailored to each customer with individualized subscription plans.
- the invention has been described as applicable to an implementation anticipating Internet based ordering and mail or other long-distance delivery of the items, where the special advantages of the method are very attractive.
- the same invention may be applied in a more conventional video, games, or music rental-store setting, where subscription customers may be allowed rentals of a specified number of movies, games, or music selections at any time, and/or in one subscription period, without rental return due dates, in exchange for a periodic rental subscription fee.
Abstract
Description
- This application is a continuation-in-part of and claims priority to U.S. patent application Ser. No. 09/884,816 (Attorney Docket No. 56055-0013) filed on Jun. 18, 2001, the entire contents of which are incorporated herein by reference in their entirety for all purposes, which is a continuation-in-part of U.S. patent application Ser. No. 09/561,041, now issued as U.S. Pat. No. 6,584,450 (Attorney Docket No. 56055-0011), filed on Apr. 28, 2000 and also claims priority to U.S. Provisional Patent Application No. 60/212,193 (Attorney Docket No. 56055-0012) filed on Jun. 16, 2000 and U.S. Provisional Patent Application No. 60/244,793 (Attorney Docket No. 56055-0014) filed on Oct. 31, 2000.
- The present invention relates to inventory rental, and more specifically, to an approach for estimating user ratings of items.
- Conventional inventory rental models are typically based upon renting items for fixed rental periods and charging late fees for keeping rented items beyond a specified return date. As used herein, the term “item” refers to any item of commerce. These types of inventory models suffer from several significant limitations. First, conventional rental models require customers to make the decision of what items to rent at substantially the same time as the decision of when to rent the items. An example that illustrates this limitation is a video rental business. Customers go to a video rental store and select particular movies to rent at that time. The customers take the movies home and must return them by a particular due date or be charged a late fee. In this situation, the customers cannot decide what movies to rent before actually renting them. The customers may have a particular movie in mind, but there is no guarantee that the video rental store has the particular movie in stock. Moreover, due dates are inconvenient for customers, particularly for “new release” movies that are generally due back the next day.
- Given the current demand for inventory rental and the limitations in the prior approaches, an approach for renting items to customers that does not suffer from limitations associated with conventional inventory rental models is highly desirable. In particular, an approach for renting inventory items to customers that allows separation of customers' decisions of what items to rent from when to rent the items is highly desirable.
- There is a further need for an approach for renting items to customers on a continuous basis that avoids the use of fixed due dates or rental “windows” appurtenant to conventional rental models.
- There is yet a further need for an approach for renting movies, games and music to customers that is more convenient and flexible to customers than conventional approaches.
- There is also a need for an approach for estimating how a user would rate an item that the user has not yet rated.
- Embodiments of the invention are illustrated by way of example, and not by way of limitation, in the figures of the accompanying drawings and in which like reference numerals refer to similar elements and in which:
-
FIG. 1 is a diagram depicting an approach for renting items to customers according to an embodiment. -
FIG. 2 is a flow diagram depicting an approach for renting items to customers according to an embodiment. -
FIG. 3 is a flow diagram depicting a “Max Out” approach for renting items to customers according to an embodiment. -
FIG. 4 is a flow diagram depicting a “Max Turns” approach for renting items to customers according to an embodiment. -
FIG. 5 is a diagram depicting an approach for renting audio/video items to customers over the Internet according to an embodiment. -
FIG. 6 is a flow diagram illustrating an approach for renting audio/video items to customers over the Internet using both “Max Out” and “Max Turns” according to an embodiment. -
FIG. 7 is a flow diagram that depicts an approach for estimating how a user would rate an item according to an embodiment of the invention. -
FIG. 8 is a block diagram that depicts users' ratings of items arranged in an array. -
FIG. 9 is a block diagram that depicts an arrangement for recommending items according to an embodiment of the invention. -
FIG. 10 is a block diagram that depicts another arrangement for recommending items according to an embodiment of the invention. -
FIG. 11 is a block diagram of a computer system upon which embodiments of the invention may be implemented. - In the following description, for the purposes of explanation, specific details are set forth in order to provide a thorough understanding of the invention. However, it will be apparent that the invention may be practiced without these specific details. In other instances, well-known structures and devices are depicted in block diagram form in order to avoid unnecessarily obscuring the invention.
- Various aspects and features of example embodiments of the invention are described in more detail hereinafter in the following sections:
-
- I. FUNCTIONAL OVERVIEW
- II. ITEM SELECTION CRITERIA
- III. ITEM DELIVERY
- IV. “MAX OUT”
- V. “MAX TURNS”
- VI. INVENTORY MANAGEMENT
- VII. ITEM RECOMMENDATION
- VIII. IMPLEMENTATION MECHANISMS
I. Functional Overview
-
FIG. 1 is a block diagram 100 that illustrates an approach for renting items to customers according to various embodiments described herein. As used herein, the term “items” refers to any commercial goods that can be rented to customers. Examples of items include movies, music and games stored on a non-volatile memory such as a tape, other magnetic medium, optical medium, read-only memory or the like, and the invention is not limited to any particular type of item. In general, the decision of what items to rent is separated from the decision of when to rent the items. Customers may specify what items to rent using one or more item selection criteria separate from deciding when to receive the specified items. Furthermore, customers are not constrained by conventional rental “windows” and instead can have continuous, serialized rental of items. - According to one embodiment, a
customer 102 provides one or more item selection criteria to aprovider 104 over alink 106.Link 106 may be any medium for transferring data betweencustomer 102 andprovider 104 and the invention is not limited to any particular medium. Examples oflink 106 include, without limitation, a network such as a LAN, WAN or the Internet, a telecommunications link, a wire or optical link or a wireless connection. - The item selection criteria indicate items that
customer 102 desires to rent fromprovider 104. In response to receiving the item selection criteria fromcustomer 102,provider 104 provides the items indicated by the item selection criteria tocustomer 102 over adelivery channel 108.Delivery channel 108 may be implemented by any mechanism or medium that provides for the transfer of items fromprovider 104 tocustomer 102 and the invention is not limited to any particular type of delivery channel. Examples ofdelivery channel 108 include, without limitation, mail delivery, courier delivery or delivery using a delivery agent.Provider 104 may be centralized or distributed depending upon the requirements of a particular application. - According to an embodiment, a “Max Out” approach allows up to a specified number of items to be rented simultaneously to
customer 102 byprovider 104. According to another embodiment, a “Max Turns” approach allows up to a specified number of item exchanges to occur during a specified period of time. The “Max Out” and “Max Turns” approaches may be used together or separately with a variety of subscription methodologies. - The approach just described for renting items to customers is now described with reference to a flow diagram 200 of
FIG. 2 . After starting instep 202, instep 204,customer 102 creates item selection criteria. Instep 206,customer 102 provides the item selection criteria toprovider 104. Instep 208, in response toprovider 104 receiving the item selection criteria fromcustomer 102,provider 104 provides one or more items indicated by the item selection criteria tocustomer 102. The process is complete instep 210. - II. Item Selection Criterial
- The one or more item selection criteria provided by
customer 102 toprovider 104 indicate the particular items thatcustomer 102 desires to rent fromprovider 104. Thus, the item selection criteria define a customer-specific order queue that is fulfilled byprovider 104. According to one embodiment, the item selection criteria specify attributes of items to be provided byprovider 104 tocustomer 102. Item selection criteria may specify any type of item attributes and the invention is not limited to particular item attributes. Examples of item attributes include, without limitation, identifier attributes, type attributes and cost attributes. Item selection criteria may be changed at any time to reflect changes in items that customers desire to rent from a provider. - III. Item Delivery
- According to one embodiment, items are delivered by
provider 104 tocustomer 102 overdelivery channel 108 based upon item delivery criteria. More specifically, the delivery of items fromprovider 104 tocustomer 102 is triggered by item delivery criteria being satisfied. The item delivery criteria may include a wide range of criteria and the invention is not limited to any particular item delivery criteria. Examples of item delivery criteria include, without limitation, customer request/notification, customer notification that an item is being returned, customer return of an item, the occurrence of a specified date, the elapsing of a specified period of time or a customer payment. - The item delivery criteria may be specified by
customer 102 toprovider 104 or negotiated bycustomer 102 andprovider 104 as part of a subscription service. For example, a particular subscription service may include item delivery criteria that specifies that a particular number of items are to be delivered monthly. As another example, item delivery criteria may specify that an initial set of items is to be delivered byprovider 104 tocustomer 102 upon initiation of a subscription service and that additional items are to be delivered tocustomer 102 upon return of items toprovider 104. Item delivery criteria may be applied uniformly to all items to be delivered to a customer, or may be item specific. For example, item delivery criteria may specify a particular date, i.e., the third Wednesday of every month, for all item deliveries. Alternatively, separate item delivery dates may be assigned to each item. - IV. “Max Out”
- According to one embodiment, a “Max Out” approach is used to manage the number of items that may be simultaneously rented to customers. According to the “Max Out” approach, up to a specified number of items may be rented simultaneously to a customer. Thus, the “Max Out” approach establishes the size of an inventory of items that may be maintained by customers. The specified number of items may be specific to each customer or may be common to one or more customers. In the present example, if the specified number of items is three, then up to three items may be rented simultaneously by
provider 104 tocustomer 102. If the specified number of items are currently rented tocustomer 102 and the specified item delivery criteria triggers the delivery of one or more additional items, then those items are not delivered until one or more items are returned bycustomer 102 toprovider 104. - According to one embodiment, in situations where the specified number of items are currently rented to
customer 102 and the specified item delivery criteria triggers the delivery of one or more additional items, then the one or more additional items are delivered tocustomer 102 andcustomer 102 and a surcharge is appliedcustomer 102. The specified number of items may then be increased thereafter to reflect the additional items delivered tocustomer 102 and increase the size of the inventory maintained bycustomer 102. Alternatively, the specified number of items may remain the same and number of items maintained bycustomer 102 returned to the prior level after items are returned toprovider 104 bycustomer 102. When used in conjunction with the “Max Turns” approach described hereinafter, the specified number of items may be unlimited. - The “Max Out” approach for managing the number of items that may be simultaneously rented to customers is now described with reference to a flow diagram 300 of
FIG. 3 . After starting instep 302, instep 304, one or more initial items are delivered tocustomer 102 to establish the inventory maintained bycustomer 102. Note that an initial delivery of items is not required and according to one embodiment, the inventory ofcustomer 102 is incrementally established over time. - In
step 306, a determination is made whether the item delivery criteria have been satisfied. If not, then the determination continues to be made until the item delivery criteria are satisfied. As described previously herein, the delivery criteria may include customer notification generally, customer notification that an item is being returned, the actual return of an item, the occurrence of a specific date, or that a specified amount of time has elapsed. - Once the item delivery criteria are satisfied, then in
step 308, a determination is made whether the specified number of items have been delivered. If not, then control returns to step 304 and one or more additional items are delivered byprovider 104 tocustomer 102. If however, instep 308, the specified number of items have been delivered, then instep 310, a determination is made whether the specified number of items, i.e., the “Max Out” limit, is to be overridden. As previously described, the specified number of items may be overridden by increasing the specified number of items, i.e., the “Max Out” limit, to allow additional items to be delivered tocustomer 102 and charging a fee tocustomer 102. Alternatively, the specified number of items is not changed and a surcharge applied tocustomer 102. This process continues for the duration of the subscription and is then complete instep 312. - V. “Max Turns”
- According to one embodiment, a “Max Turns” approach is used to rent items to customers. According to the “Max Turns” approach, up to a specified number of item exchanges may be performed during a specified period of time. For example, referring to
FIG. 1 , suppose thatprovider 104 agrees to rent items tocustomer 102 with a “Max Turns” limit of three items per month. This means thatcustomer 102 may make up to three item exchanges per month. This approach may be implemented independent of the number of items that a customer may have rented at any given time under the “Max Out” approach. The approach is also independent of the particular item delivery criteria used. - According to one embodiment, the “Max Turns” approach is implemented in combination with the “Max Out” approach to rent items to customers. In this situation, up to a specified number of total items are simultaneously rented to
customer 102 and up to a specified number of item exchanges may be made during a specified period of time. Thus, using the “Max Out” and the “Max Turns” approaches together essentially establishes a personal item inventory forcustomer 102 based upon the “Max Out” limit that may be periodically refreshed based upon the “Max Turns” limit selected. - In some situations,
customer 102 may wish to exchange more than the specified number of items during a specified period. According to one embodiment, in this situation,provider 104 agrees to rent additional items above the specified number tocustomer 102 and to chargecustomer 102 for the additional items. For example, suppose thatprovider 104 agrees to rent items tocustomer 102 with up to three item turns (exchanges) per month. If, in a particular month,customer 102 requires two additional turns, then the two additional items are provided tocustomer 102 and a surcharge is applied tocustomer 102 for the additional two items. - In other situations,
customer 102 may not use all of its allotted turns during a specified period. According to one embodiment, customers lose unused turns during a subscription period. For example, ifcustomer 102 has a “Max Turns” limit of four item exchanges per month and only makes two item exchanges in a particular month, then the two unused exchanges are lost and cannot be used. At the start of the next month,customer 102 would be entitled to four new item exchanges. - According to another embodiment, customers are allowed to carry over unused turns to subsequent subscription periods. For example, if
customer 102 has a “Max Turns” limit of four item exchanges per month and only makes two item exchanges in a particular month, then the two unused exchanges are lost and cannot be used. At the start of the next month,customer 102 would be entitled to six new item exchanges, two from the prior month and four for the current month. - The “Max Turns” approach for renting items to customers is now described with reference to a flow diagram 400 of
FIG. 4 . After starting instep 402, instep 404,customer 102 andprovider 104 agree upon the terms of the “Max Turns” agreement. Specifically,customer 102 andprovider 104 agree at least upon the maximum number of turns that are allowed in a specified period of time. - In
step 406, in response to one or more item delivery criteria being satisfied,provider 104 provides one or more items tocustomer 102 overdelivery channel 108. Any type of item delivery criteria may be used with the “Max Turns” approach and the invention is not limited to any particular delivery criteria. For example, the initial one or more items may be delivered tocustomer 102 in response to a subscription payment made bycustomer 102 toprovider 104, the initiation of a specified subscription period, or by request ofcustomer 102 for the initial-rental items. The number of initial one or more items must not exceed the terms of the “Max Turns”agreement. - In
step 408, in response to one or more delivery criteria being satisfied, a determination is made whether additional items can be provided tocustomer 102 within the terms of the “Max Turns” agreement. For example, if the number of items rented to customer in the current subscription period is less than the agreed-upon “Max Turns,” then additional items can be rented tocustomer 102 within the terms of the “Max Turns” agreement. In this situation, this determination may be made in response tocustomer 102 returning one or more items toprovider 104, or bycustomer 102 requesting additional items. - If, in
step 408, a determination is made that additional items can be rented tocustomer 102 within the terms of the “Max Turns” agreement, then control returns to step 406 where one or more additional items are rented tocustomer 102. If however, instep 408, a determination is made that additional items cannot be rented tocustomer 102 within the-terms of the “Max Turns” agreement, then instep 410, a determination is made whether to override the current agreement terms. If so, then instep 412, the agreement terms are changed to allow for a larger number of terms andcustomer 102 is charged accordingly, or the terms are left unchanged and a surcharge is applied for the additional items to be delivered. Control then returns to step 406, where one or more additional items are delivered tocustomer 102. - If in
step 410, a determination is made that the current agreement is not to be overridden, then instep 414, no items are delivered tocustomer 102 until the next subscription period. For example, the request for additional items may be received at the end of a subscription period and instead of renting the additional items immediately, they are instead delivered during the subsequent subscription period. Control then returns to step 406 where one or more additional items are rented to customer or the process is complete instep 416. - The approach for renting items described herein is now described in the context of renting to customers audio/video (A/V) items, such as movies, games and music, stored on various media.
FIG. 5 is a diagram 500 that depicts a set ofcustomers 502 that desire to rent A/V items from aprovider 504.Customers 502 communicate withprovider 504 overlinks 506, the global packet-switched network referred to as the “Internet,” and alink 510. -
Links customers 502 and theInternet 508 and between theInternet 508 andprovider 504, respectively, and the invention is not limited to any particular medium. In the present example,links customers 502 are configured with generic Internet web browsers.Links - In accordance with an embodiment,
customers 502 enter into a rental agreement withprovider 504 to rent audio/video (A/V)items 512 fromprovider 504 according to the “Max Out” and/or “Max Turns” approaches described herein. The invention is not limited to any particular approach for entering into the rental agreement. For example,customers 502 andprovider 504 may enter into a rental agreement by mail, telephone or over the Internet, bycustomers 502 logging into a web site associated withprovider 504. -
Customers 502 create and provide item selection criteria toprovider 504 overlinks Internet 508. The invention is not limited to any particular approach for specifying and providing item selection criteria toprovider 504. For example, according to one embodiment,customers 502 provide item selection criteria toprovider 504 in one or more data files. According to another embodiment,customers 502 log onto a web site ofprovider 504 and use a graphical user interfaced (GUI) to specify attributes of the movies and music that customers desire to rent fromprovider 504. - The item selection attributes may include any attributes that describe, at least in part, movies, games or music that
customers 502 desire to rent. For movies, example attributes include, without limitation, title, category, director name, actor name and year of release. For games, example attributes include, without limitation, title and category. For music, example attributes include, without limitation, title, category, artist/group name and year of release.Customers 502 may identify specific movies or music by the item selection criteria, or may provide various attributes and allowprovider 504 to automatically select particular movies and music that satisfy the attributes specified. For example,customers 502 may specify item selection criteria that include horror movies released in 1999 and letprovider 504 automatically select horror movies that were release in 1999. As another example,customers 502 may specify item selection criteria that include adventure movies starring Harrison Ford.Customers 502 may also specify an order or priority for the specified item selection criteria. For example,customers 502 may specify specific movie titles and the order in which they want to receive them. As another example,customers 502 may specify that they want to receive a particular number of movies of different types. - Once
customers 502 andprovider 504 have entered into a rental agreement andcustomers 502 have provided item selection criteria toprovider 504, then A/V items 512 are rented tocustomers 502 overdelivery channels 514 in accordance with the terms of the rental agreement. Specifically, according to the “Max Out” approach described herein, an initial set of A/V items 512, such as movies, games and music, are delivered tocustomers 502 overdelivery channels 514 according to the terms of the rental agreement. Subsequent A/V items 512 are delivered whenever the specified item delivery criteria are satisfied. For example, additional A/V items 512 may be delivered upon the return of one or more A/V items 512 to provider, a request fromcustomers 502, the arrival of a particular date, e.g., a specific day of the month, or the expiration of a specified period of time, e.g., fifteen days. - In accordance with the “Max Out” approach described herein, once the maximum number of A/
V items 512 have been rented to aparticular customer 502, then no additional A/V items 512 are rented until one or more rented A/V items 512 are returned toprovider 504, or unless a surcharge is applied to theparticular customer 502. Alternatively, the rental agreement between theparticular customer 502 andprovider 504 may be modified to increase the maximum number of A/V items 512 that may be rented simultaneously to theparticular customer 502. - The rental agreement between
customers 502 andprovider 504 may also specify a maximum number of turns in combination with the “Max Out” approach. In this situation, a maximum number of turns restricts how quicklycustomers 502 may refresh their A/V item 512 inventories. For example, suppose that aparticular customer 502 agrees withprovider 504 to rent up to four movies with a maximum of four turns per month. Under this agreement, theparticular customer 502 may maintain a personal inventory of up to four movies and rent four new movies per month. Thus, theparticular customer 502 can completely “replace” his personal inventory once per month. If theparticular customer 502 agreed to a maximum of up to eight turns per month, then theparticular customer 502 would be able to completely replace his personal inventory twice per month. -
Provider 504 is illustrated as a single entity for purposes of explanation only.Provider 504 may be centralized or distributed depending upon the requirements of a particular application. For example,provider 504 may be a central warehouse from which all A/V items 512 are provided. Alternatively,provider 504 may be implemented by a network of distributed warehouses. -
FIG. 6 is a flow diagram that illustrates an approach for renting A/V items 512, e.g., movies, to customers over a communications network such as the Internet using both “Max Out” and “Max Turns” according to an embodiment. Referring also toFIG. 5 , after starting instep 602, instep 604, acustomer 502 enters into a rental agreement withprovider 504. In the present example,customer 502 uses a generic web browser to access an Internet web site associated withprovider 504 and enter into a rental agreement that specifies thatcustomer 502 may maintain a personal inventory of four movies (“Max Out” of four) and receive up to four new movies per month (“Max Turns” of four). Furthermore, the rental agreement specifies that new movies will be delivered upon return of a rented movie fromcustomer 502, i.e., the delivery criteria is a return of a movie by the customer. - In
step 606,customer 502 creates and provides movie selection criteria toprovider 504 that indicates movies thatcustomer 502 desires to rent. For example, the movie selection criteria may specify particular movie titles thatcustomer 502 desires to rent. The movie selection criteria may also specify an order or priority in whichcustomer 502 wishes to rent the movies. Instead of identifying particular movie titles, the movie selection criteria may specify movie preferences forcustomer 502, e.g., types of movies, directors, actors, or any other movie preferences or attributes. In this situation,provider 504 automatically selects particular titles that satisfy the movie selection criteria. For example, the movie selection criteria may specify a preference for action movies starring a particular actor, with a preference for “new release” movies.Provider 504 attempt to provide movies tocustomer 502 that best satisfy the preferences indicated by the movie selection criteria. - In
step 608, one or moreinitial movies 512 are delivered tocustomer 502 overdelivery channel 514. The one or moreinitial movies 512 may be delivered tocustomer 502 via mail, courier, delivery agent or any other suitable means negotiated betweencustomer 502 and provider and the invention is not limited to any particular type of delivery mechanism. For purposes of explanation only, is presumed in the present example that movies are mailed betweencustomer 502 andprovider 504. - The one or more
initial movies 512 establish the personal movie inventory ofcustomer 502.Customer 502 may choose to receive any number of movies up to the “Max Out” limit of four movies. Typically,customer 502 will choose to initially receive four movies in the initial delivery. - Once the one or more
initial movies 512 have been mailed tocustomer 502, then instep 610, a determination is made whether anymovies 512 have been returned bycustomer 502 to trigger another movie delivery. In the present example, the delivery of additional movies is triggered by the receipt, e.g., via mail, of one or more movies fromcustomer 502. In the situation wherecustomer 502 elects to not receive the maximum number ofmovies 512 in the initial delivery, then the delivery ofadditional movies 512 may also be triggered by a request fromcustomer 502 foradditional movies 512. For example,customer 502 may notifyprovider 504 via telephone, email or by accessing the web site associated withprovider 504. - If, in
step 610, a determination is made that one ormore movies 512 were received fromcustomer 502, then instep 612, a determination is made whether the maximum number of turns (“Max Turns”) limit has been reached for the current cycle. In the present example, a determination is made whether four or more movies have been mailed in the current month. If not, then control returns to step 608, where one or moreadditional movies 512 are mailed tocustomer 502 viadelivery channel 514 up to the “Max Out” limit of four. - If, in
step 612, a determination is made that the “Max Turns” limit has been met for the current cycle, i.e., in the present example, fourmovies 512 have been mailed tocustomer 502 in the current month, then in step 614 a determination is made whether to override the current “Max Turns” limit. If so, then instep 616, a surcharge is applied tocustomer 502 and control returns to step 608 where theadditional movies 514 are mailed tocustomer 502. If not, then instep 618, a determination is made whether to continue the subscription service. If so, then no additional movies are mailed tocustomer 502 during the current cycle, e.g., the current month, and the control returns to step 610. If, instep 618, a determination is made that service is not to be continued, then the process is complete instep 620. - In some situations,
customer 502 may desire to increase or decrease the size of customer's 502 personal movie inventory by changing the current “Max Out” limit. According to one embodiment,customer 502 notifiesprovider 504, e.g., by telephone, mail, email or by accessing the web site associated withprovider 504, thatcustomer 502 wishes to change the “Max Out” limit. The movie rental agreement betweencustomer 502 andprovider 504 is then modified to reflect the change of the “Max Out” limit. In the situation where the “Max Out” limit is increased, thenadditional movies 512 may be immediately mailed tocustomer 502. - VI. Inventory Management
- The approach described herein for renting items to customers provides superior inventory management to prior approaches. Specifically, the use of item selection criteria provides for efficient inventory management by allowing the greatest number of items to be rented at any given time. Moreover, the greatest number of customers are provided with their most preferred items. For example, customers may specify priorities for the items indicated by the item selection criteria. Thus, if a particular customer's first choice is not available, or already rented, then the item having the next highest priority can be rented to the particular customer. According to one embodiment, customers may indicate items that are not yet available for rent. Then, the items are delivered to customers when they become available.
- For example, referring again to
FIG. 5 , suppose that aparticular customer 502 desires to rent an as-yet-unreleased movie entitled “ABC.” Theparticular customer 502 indicates this movie toprovider 504 by the item selection criteria. Since the movie ABC is not yet available, it cannot be delivered to theparticular customer 502. However, when the movie ABC does become available, it can be shipped immediately to theparticular customer 502, as well asother customers 502 who may have also requested the movie. This allowsprovider 504 to maximize the number of items rented while ensuring thatcustomers 502 are able to rent the highest priority items that they requested. - According to another embodiment, as yet unknown items may also be rented by specifying attributes of the unknown items. For example, the
particular customer 502 may request to rent the next new movie of a particular director, for which the exact name is unknown to the particular customer. As another example, theparticular customer 502 may request to rent the next album of a particular group that is currently in process and does not yet have a title. - VII. Item Recommendation
- A. Overview
- In some situations, it is desirable to know how a user would rate an item. For example, in the context of recommending items to users, it is useful to know how a user would rate an item, so that a decision can be made to either recommend or not recommend the item to the user. A recommendation is more valuable to the user if it is likely that the user would rate the item highly because the user is more likely to be happy with the recommendation. An approach is provided for estimating how a user would rate an item. The approach generally involves estimating how a user would rate an item based upon similarities between items and similarities of ratings given to items by both the user and one or more other users. The approach is applicable to any type of item, for example rental items such as movies and games, and the invention is not limited to any particular type of item. The approach is applicable to a wide variety of contexts and may be used, for example in a recommendation engine to recommend items to users.
-
FIG. 7 is a flow diagram 700 that depicts an approach for estimating how a user would rate a particular item according to an embodiment of the invention. Instep 702, one or more items from a plurality of items are identified that have ratings similar to the particular item. Instep 704, one or more other users are identified that have given ratings to the one or more other items that are substantially similar to ratings given by the user to the one or more other items, and have also rated the particular item. Instep 706, an estimation of how the user would rate the particular item is generated based upon ratings given by the one or more other users for the particular item. Each of these steps is described in more detail hereinafter. -
FIG. 8 is a block diagram that depicts users' ratings of items arranged in anarray 800.Array 800 includes columns ofitems 802 and rows of users 804. Each cell may be empty or contain data indicating that the user has not provided a rating for the item. For example, a specified value, such as a null value or other value, may be used to indicate that a user has not provided a rating for an item. Alternatively, each cell may contain data that represents a user rating of an item. Different types of data and values may be used to indicate a user rating of an item, depending upon the requirements of a particular application, and the invention is not limited to any particular data types or values. For example, ratings may be represented by numerical data, e.g., in a range of 1 to 5 that corresponds to a 5-star rating scale or binary data that corresponds to a favorable or unfavorable rating. Embodiments of the invention are described herein in the context of users' ratings inarray 800 for explanation purposes only and the invention is not limited to users' ratings being maintained in any particular format or data structure. Furthermore, the relative size of theitems 802 and users 804 dimensions depicted inFIG. 8 is not meant to imply that there must be more users 804 thanitems 802. Embodiments of the invention are described herein in the context of generating an estimated rating for aparticular item 806, fromitems 802, and aparticular user 808, from users 804. The estimated rating for theparticular item 806 andparticular user 808 is indicated inFIG. 8 byreference numeral 810. - B. Identifying Other Items that have Ratings Similar to the Particular Item
- A set of
similar items 812 that have ratings similar to theparticular item 806 is identified.Similar items 812 are depicted inFIG. 8 as being located adjacent to theparticular item 806 inarray 800 for explanation purposes only.Similar items 812 may include any ofitems 802. A variety of techniques may be used to determinesimilar items 812, depending upon the requirements of a particular application, and the invention is not limited to any particular technique. - According to one embodiment of the invention,
similar items 812 are items fromitems 802 that satisfy item similarity criteria with respect toparticular item 806. Thus, other items that satisfy the item similarity criteria with respect to the particular item are considered to be similar to the particular item. Similarity is not necessarily symmetrical. Thus, although an item A may be considered similar toparticular item 806 and therefore included insimilar items 812 forparticular item 806,particular item 806 may, or may not be, sufficiently similar to item A to be considered similar to item A and included in a set of similar items for item A. The number of items insimilar items 812 may be limited to a specified number to reduce the amount of computational resources and time required to generate an estimated rating forparticular item 806. For example, the specified number of items insimilar items 812 may represent the first specified number of items that satisfied the item similarity criteria forparticular item 806, or may represent the specified number of items that satisfied the item similarity criteria forparticular item 806 with similarity above some specified level or threshold. - The item similarity criteria may include a wide variety of criteria, depending upon the requirements of a particular implementation, and the invention is not limited to any particular item similarity criteria. One example of item similarity criteria is a statistical correlation threshold. Thus, two items that satisfy the statistical correlation threshold are considered to be similar to each other. A wide variety of statistical measures may be used for this purpose, depending upon the requirements of a particular implementation, and the invention is not limited to any particular statistical measure of similarity. For example, a Pearson Correlation Coefficient may be used as the statistical measure of similarity as shown in equation (1) below:
where Xi are the ratings for one item or user and Yi are the ratings for the other item or user, and the summation is applied across all co-ratings. As another example, a vector cosine distance may be used as the statistical measure of similarity as shown in equation (2) below:
where Xi and Yi are the ratings for one item or user. - Table I below contains example user ratings for items X and Y, the product of the user ratings for X and Y, the square of the user ratings for X and Y and the summations of these values:+
TABLE I X Y X · Y X**2 Y**2 1 2 2 1 4 4 4 1 4 16 1 5 4 20 25 16 3 3 9 9 9 3 5 4 20 25 16 1 4 4 16 16 16 4 5 20 16 25 Summations 26 23 91 108 87 - Note that the summations of the X and Y columns include only ratings for users who rated both item X and Y.
- Another example of the item similarity criteria is a specified number of co-ratings between items. A co-rating exists for two items when a user rates both items. Thus, the item similarity criteria may specify a minimum number of co-ratings that must be shared between two items for the two items to be considered similar items. Yet another example of the item similarity criteria is a specified fraction of co-ratings between items divided by the number of ratings of either item. Thus, the item similarity criteria may specify a minimum fraction of co-ratings between two items divided by the number of ratings of either item that must be satisfied for the two items to be considered similar items. If the total population of users is U, the number of users who rated item A is Ra, the number of users who rated item B is Rb, and if the probability of a user rating item A is independent of the probability of the user rating item B, then the expected number of co-ratings E(Rab) is given by:
If the actual number of co-raters of both Item A and Item B Rab is much larger than E(Rab), it is likely that Item A and Item B are related in some way. On the other hand, if Rab is much smaller than E(Rab), it is likely that Item A and Item B are dissimilar in a way such that raters of Item A are unlikely to encounter Item B, or unlikely to have an opinion about Item B and so are less likely to rate Item B (and vice-versa). Either of these pieces of information may be used as additional similarity criteria. Note: if one of Item A or Item B is newer than the other (in other words, did not exist at the time the other Item was popular or well-known), then it is very likely that the total number of co-ratings of the Items will be much smaller than would be expected by the basic formula above. As a variation, the pool of ratings may be subdivided into slices by time—for example, months—and the co-raters fraction calculation described above may be applied to in each successive (month) slice. - In some situations, using the Pearson Correlation Coefficient to measure similarity between two items can make two dissimilar items appear similar. This occurs where the underlying rating distributions of two items make the co-ratings of the two items appear to be correlated. For example, two items may have a high number of high ratings. Although the items are unrelated, a grouping of rating pairs in both items among the co-raters, i.e., users who rated both items, makes them appear related.
- According to one embodiment of the invention, the ratings distributions of the two items are “factored out.” This is performed by taking the percentage of each rating for each item, and multiplying it out in a two-dimensional array to obtain expected percentages for every possible pair of ratings (r1, r2) for the two items. Multiplying each of the array values by the total number of co-raters produces the expected count of co-raters for each rating pair, if the items are unrelated. The actual counts of co-ratings in each of the array cells are compared to see if they are more or less than expected. Higher than expected values in areas corresponding to similarity, e.g. a user that rated both items high or both item slow, indicates that the items are similar. Higher than expected co-rater counts in areas where a user liked one item and disliked another indicates that the items are not positively related. This use of the underlying ratings distribution for all raters provides a better understanding of values for co-raters.
- The descriptions above have identified several different similarity criteria, which may be combined to determine overall similarity. For example, in one embodiment, the Pearson Correlation Coefficient is required to be greater than a threshold (e.g. 0.3), and also the fraction of co-raters Rab must be at least some fraction (e.g. one half) the expected number of co-raters E(Rab), and the absolute number of co-raters must be at least some minimum number (e.g. 200).
- The methods described above show how to calculate the similarity between two items in order to find a set of similar items. The exact same methods may be applied in order to calculate the similarity between two users in order to find a set of similar users.
- In some circumstances, there may be insufficient rating data to determine a set of items similar to the particular item. For example, if the particular item is new, there may be very little data allowing correlations to be calculated. In this case, other semantic information may be used to help determine similar items.
- In one embodiment, items are additionally classified by semantic feature values. For example, if the items are movies, the semantic features might be genre (or category), director, lead actors/actresses, other actors/actresses, etc., and the feature values are the genre, director, or actor names. In a music domain, the features might be styles, soloists, musicians, and lyricists. Each type of feature is given an importance score, which may be positive, negative, or zero, and two items may be compared by counting the number of feature matches and summing their importance scores. For example, if a movie director has an importance score of 3, and a lead actor has an importance score of 1, then a pair of movies sharing one director and one lead actor would score a similarity of 4, while a different pair of movies sharing two lead actors would have a similarity score of 2. The absence of matches may be accorded a negative score (for example, the absence of matching directors might score the pairing −0.1 points). The set of similar items is the specified number of items whose similarity score with the particular item is largest, or the set of items whose similarity score is above a specified threshold.
- In another embodiment, tools may be provided for domain experts to specify item similarity by non-machine-implemented means. Once additional ratings data becomes available, the similar items specified by domain experts may be supplemented or discarded.
- C. Identifying Other Users that have Given Ratings for the Other Items Similar to the Particular User
- Once
similar items 812 have been determined, one or more users from users 804 that have given ratings tosimilar items 812 that are similar to ratings given by theparticular user 808 tosimilar items 812 are identified. All of users 804 may be evaluated for this purpose. There may be a significant and unacceptable computational cost, however, in evaluating all of users 804 in some applications where the number of users 804 is large. Furthermore, some users may be less useful than others because they have very little ratings data or have ratings data with poor statistical characteristics, as described in more detail hereinafter. - According to one embodiment of the invention, a subset of users 804, referred to herein as
reference users 814, is determined based upon specified user selection criteria.Reference users 814 are users from all users 804 that satisfy the specified user selection criteria. A wide variety of user selection criteria may be used, depending upon the requirements of a particular application, and the invention is not limited to any particular user selection criteria. According to one embodiment of the invention, the user selection criteria include a minimum number of ratings. Thus, users from users 804 that do not have at least the minimum number of ratings are not included inreference users 814. The user selection criteria may include a maximum number of ratings. Thus, users from users 804 that have more than the maximum number of ratings are not included inreference users 814. - According to another embodiment of the invention, the user selection criteria specify an average rating range. Thus, users from users 804 with an average rating that is outside the specified range are not included in
reference users 814. For example, suppose that ratings are provided on a numerical scale from one to five. Suppose further that the user selection criteria specify that the acceptable average rating range is two to four. In this situation, users 804 with an average rating that is less than two or greater than four are not included inreference users 814. - According to another embodiment of the invention, the user selection criteria specify that user ratings must conform to within a specified tolerance of a Normal distribution. Thus, users from users 804 with a rating distribution that does not approximate a Normal distribution within a specified tolerance are not included in
reference users 814. For example, suppose that ratings are on a whole number numerical scale of one to five. A user that has more one ratings than two ratings, more two ratings than three ratings, more five ratings than four ratings, or more four ratings than three ratings is not included inreference users 814. The specified tolerance may vary, depending upon the requirements of a particular application. For example, some applications may use a specified tolerance that allows a user inreference users 814 to have a slightly greater number of two ratings than three ratings, and an equal number of two ratings and one ratings, but not more one ratings than three ratings. - According to another embodiment of the invention, the user selection criteria specify a range of acceptable standard deviations of user ratings. Thus, users from users 804 with a standard deviation of ratings that is not within the range of acceptable range of standard deviations of ratings are not included in
reference users 814. For example, the user selection criteria might specify a standard deviation range of between 10% and 20%. Thus, a user having a standard deviation of ratings of less than 10% or 20% or greater is not included inreference users 814. - D. Generate Estimation of how the Particular User Would Rate the Particular Item
- Generating an estimation of how
particular user 808 would rateparticular item 806 generally involves identifying which users 804 have ratedsimilar items 812 items similarly toparticular user 808, and then generating the estimation based upon how those users 804 rated theparticular item 806. In the situation wherereference users 814 are identified from users 804 as described herein, then the approach generally involves identifying whichreference users 814 have ratedsimilar items 812 items similarly toparticular user 808, and then generating the estimation based upon how thosereference users 814 rated theparticular item 806. Embodiments of generating an estimation of howparticular user 808 would rateparticular item 806 are described herein in the context of usingreference users 814, although the invention is not limited to this context. - According to one embodiment of the invention, reference users' 814 ratings of
similar items 812 are compared to the particular user's 808 ratings ofsimilar items 812 to identifyreference users 814 that are the most similar toparticular user 808. InFIG. 8 ,reference users 814 that have provided ratings forsimilar items 812 that are most similar to the ratings forsimilar items 812 provided byparticular user 808 are referred to hereinafter assimilar reference users 816. InFIG. 8 , the horizontal lines depicted acrosssimilar items 812 for thesimilar reference users 816 are provided for illustration purposes only and are not meant to imply that thesimilar reference users 816 have necessarily provided ratings for allsimilar items 812. Similarly, it is not necessary for theparticular user 808 to have rated allsimilar items 812. - According to one embodiment of the invention, rating similarity criteria are used to identify the
similar reference users 816. The rating similarity criteria may be similar to the item similarity criteria described herein. For example, thesimilar reference users 816 may be determined based upon determining statistical correlations between the ratings ofsimilar items 812 provided byreference users 814 and theparticular user 808 and then selecting thesimilar reference users 816 based upon thereference users 814 that best correlate to theparticular user 808. The similarity betweensimilar items 812 and theparticular item 806 may also be considered in identifying thesimilar reference users 816. More specifically, ratings ofsimilar items 812 that are more similar to theparticular item 806 may be given more weight in determining thesimilar reference users 816 fromreference users 814. - Once the
similar reference users 816 are identified, the estimation of how theparticular user 808 would rate theparticular item 806 is generated based upon the ratings of theparticular item 806 made by thesimilar reference users 816. The degree of similarity between thesimilar reference users 816 and theparticular user 808 may vary based upon a variety of factors including, for example, the number ofsimilar items 812 that both theparticular user 808 and thesimilar reference users 816 have rated and how theparticular user 808 and thesimilar reference users 816 rated thesimilar items 812. The estimation may be generated in different ways, depending upon the similarity between the ratings ofsimilar items 812 provided by thesimilar reference users 816 and theparticular user 808. For example, there may be situations where the ratings provided by the one or moresimilar reference users 816 forsimilar items 812 are identical to the ratings provided by theparticular user 808 for the samesimilar items 812. This may occur, for example, where the number ofsimilar items 812 rated by both thesimilar reference users 816 and theparticular user 808 is relatively small. In this situation, the estimated rating of theparticular item 806 by theparticular user 808 may be the average rating made by the one or moresimilar reference users 816 for theparticular item 806. - In situations where the ratings for
similar items 812 by thesimilar reference users 816 and theparticular user 808 are not identical, then the estimated rating of theparticular item 806 by theparticular user 808 may be the weighted linear sum of the ratings made by the one or moresimilar reference users 816 for theparticular item 806. The weighting applied varies based upon the similarity between eachsimilar reference users 816 and theparticular user 808. According to one embodiment of the invention, the weighted linear sum of the ratings forparticular item 806 is the sum of the products of the weights and the similar reference users' 816 ratings ofparticular item 806, divided by the sum of the weights. Thus, the ratings ofparticular item 806 provided by thesimilar reference users 816 that are most similar to theparticular user 808 are given more weight than the ratings ofparticular item 806 provided by thesimilar reference users 816 that are less similar to theparticular user 808. For example, suppose that thesimilar reference users 816 include users A and B that are determined to have similarities to theparticular user 808 of 0.8 and 0.5 (on a scale of 0 to 1), respectively. Furthermore, users A and B have given ratings forparticular item 806 of 5 and 3 (on a scale of 1 to 5), respectively. The average rating ofparticular item 806 by users A and B is 4 ((5+3)/2). The weighted linear sum of the ratings forparticular item 806 by users A and B is 4.23 ((5*0.8+3*0.5)/1.3). The weighted linear sum of the ratings is greater than the average of the ratings in this situation because the higher rating of user A is given greater weight than the lower weighting of user B, by virtue of the relatively greater similarity between user A and theparticular user 808 than between user B and theparticular user 808. - Various confidence metrics may be calculated that indicate a confidence level in an estimated rating of
particular item 806 by theparticular user 808. According to one embodiment of the invention, a confidence metric is calculated based upon the disparity or spread between estimated ratings, where each estimated rating is based upon a different one of thesimilar reference users 816. The standard deviation of the estimated ratings may be used for this purpose. The calculation of confidence based upon the disparity between estimated ratings may weight these estimated ratings based on the strength of the similarity between thereference user 816 and theparticular user 808. A confidence metric may also be made based upon similarities between theparticular user 808 and eachsimilar reference user 816 that is used to estimate a rating. The calculation of confidence may also take into account the absolute strength of the similarity between thereference users 816 used to generate the prediction and theparticular user 808. - There may be situations where there is an insufficient (or zero) number of
similar reference users 816. In these situations, a weighted average of allreference users 814, who each share at least a rating of one ofsimilar items 812 with theparticular user 808 and who have each rated theparticular item 806, may be used to generate an estimation of how theparticular user 808 would rate theparticular item 806. According to one embodiment of the invention, the weightings are applied based upon the similarity of eachreference user 814 to theparticular user 808. Confidence metrics may also be applied to generating an estimation of how theparticular user 808 would rate theparticular item 806 based upon allreference users 814 used to generate the estimate. - In situations where there is an insufficient (or zero) number of
reference users 814, which may also occur if theparticular user 808 has an insufficient (or zero) number of ratings, an average rating of theparticular item 806 by allreference users 814 who have rated theparticular item 806 may be used as an estimate of how theparticular user 808 would rate theparticular item 806. Alternatively, an average rating of all users 804 who have rated theparticular item 806 may be used as an estimate of how theparticular user 808 would rate theparticular item 806. In these situations, a weighted average of allreference users 814, who each share at least a rating of one ofsimilar items 812 with theparticular user 808 and who have each rated theparticular item 806 may be used. - E. Recommendation Architecture
-
FIG. 9 is a block diagram that depicts anarrangement 900 for recommending items according to an embodiment of the invention.Arrangement 900 includes aWeb server 902 communicatively coupled to arecommendation system 904. As described in more detail hereinafter,recommendation system 904 is configured to process ratings data using prioritized tasks to provide improved availability and accuracy of item recommendations.Recommendation system 904 includes a recommendation engine 906, anon-volatile storage 908 and a ratings database (DB) 910. Recommendation engine 906 may be implemented by any mechanism or any combination of hardware and software for performing the functions described herein. For purposes of explanation only, recommendation engine 906 is described in the context of a software process executing in a volatile memory.Non-volatile storage 908 may be implemented by any type of non-volatile storage mechanism and the invention is not limited to any particular mechanism. Example implementations include, without limitation, non-volatile memory and disc storage.Ratings DB 910 may be implemented by any database management mechanism and the invention is not limited to any particular implementation. - Recommendation engine 906 maintains ratings and
prediction data 912 that includes users' actual ratings and estimated ratings calculated according to the algorithms described herein, or otherwise. Ratings andprediction data 912 may be maintained in a volatile memory, such as a random access memory (RAM), to reduce access times. According to one embodiment of the invention, ratings andprediction data 912 is maintained in a RAM array and is addressed using 64 bit addressing as provided by high-end computing architectures. The array is laid out such that two bytes are used for each cell, containing a rating value, an estimated rating value, a confidence value, and one or more flags indicating the presence and validity of the values. The RAM array may be considered as divided into a number of pages of perhaps 8,192 bytes each (which corresponds to the physical structure of current computers). The cells are laid out so that each user's ratings and estimated ratings span one or more pages for locality of access. - According to one embodiment of the invention, ratings and
prediction data 912 is periodically saved to adata file 914 maintained onnon-volatile storage 908. For example, low-level operating system primitives may be used to ensure that data written to the RAM array is also stored in data file 914 in a timely manner, so that recommendation engine 906 may be periodically shut down for maintenance and upgrade without losing any of the ratings andprediction data 912. Upon restarting the recommendation engine 906, data from data file 914 is mapped back (reloaded) into the RAM array, thereby restoring all the results of previous computations. In one embodiment of the invention, the memory mapping functionality of the Solaris operating system offered by Sun Microsystems, Inc., permits data file 914 to be sparsely populated with disc blocks, such that any pages in the RAM array that have no values do not occupy space onnon-volatile storage 908. This provides considerable savings in hardware cost, since much of the RAM array will never contain ratings and prediction data. Ratings data is also stored inratings DB 910, which allows data file 914 to be re-generated in the event of a failure of non-volatile storage.Ratings DB 910 may also store any data contained in the RAM array or data file 914 or any other data used byrecommendation system 904, depending upon the requirements of a particular implementation. - The use of data file 914 also allows storage of more ratings and
prediction data 912 than can be stored in the RAM array, given practical limitations in available RAM on computer architectures. According to one embodiment of the invention, the size of the RAM array is selected to be large enough to store ratings andprediction data 912 for at leastreference users 814 and active users ofrecommendation system 904. As users' status changes from active to inactive, the physical pages of storage in the RAM array allocated to their rating andprediction data 912 may be reclaimed and reused for other users just becoming active. - Recommendation engine 906 is configured to perform a wide variety of tasks. These tasks may be prioritized in one or
more task queues 916 to coordinate the activity of one or more processing threads. The one or more task queues may be stored in a volatile storage, such as RAM, or innon-volatile storage 908. The processing threads process tasks from thetask queues 916 based upon priority and the processing of a task may create additional queue requests that are added totask queues 916. For example, recommendation engine 906 is configured to process requests for estimated ratings and requests for recommendations. An estimated rating for a particular item for a particular user may be requested in situations where, for example, a user is viewing data associated with the particular item, such as a review, on a graphical user interface (GUI), and it is desirable to display data that indicates an estimate of how the particular user would rate the particular item. A request for recommendations may be made, for example, in situations where a user has requested a list of recommended items from a Web application. When a request for an estimated rating a particular item for a particular user is received fromWeb server 902, recommendation engine determines whether the requested estimated rating has previously been calculated and stored in the RAM array. If so, then the estimated rating is immediately returned toWeb server 902. If not, then a request is queued on the task queue with high priority to calculate the estimated rating (according to the algorithms described herein, or otherwise). If the compute threads complete the calculation of the estimated rating within some small threshold of time (for example, half a second) then the calculated estimated rating is returned. Otherwise, the recommendation engine 906 is configured to provide an average rating of the particular item as the estimated rating. When a request for a recommendation for a specific user is received fromWeb server 902, recommendation engine 906 identifies the set of items with the highest estimated ratings and highest confidence for that user and returns toWeb server 902 data that identifies the set of items as recommendations. If there are no confident predictions, then no recommendations may be returned. Although embodiments of the invention are described herein and depicted in the figures in the context of asingle Web server 902, this is provided for purposes of explanation only.Recommendation system 904 may be used with any number of Web servers and also other types of computing entities. - Recommendation engine 906 is also configured to receive and process rating data from
Web server 902. Upon receipt of rating data, which may include a new or changed rating, recommendation engine 906 stores the rating in the RAM array using a simple address calculation and a simple memory write. The rating is also queued in atask queue 916 to be written toratings DB 910. In addition, estimated ratings for items similar to the item associated with the received rating data may need to be updated to reflect the received rating data. Therefore, one or more tasks are added totask queues 916 to recalculate the estimated ratings for items for which the item associated with the received rating data is considered similar. According to one embodiment of the invention, these tasks are given high priority, since the updated estimated ratings may be required in a relatively short time, for example for a merchandising decision or a recommendation for a user who is presumably visiting the Website in order to provide the rating. Quality metrics may be calculated from comparisons of actual user ratings to estimated user ratings. For example, suppose that the estimatedrating 810 has been determined for theparticular item 806 and theparticular user 808. Suppose further that this estimatedrating 810 has been provided toWeb server 902, displayed for theparticular user 808, and theparticular user 808 has provided an actual rating for theparticular item 806. A quality metric may be calculated based upon a comparison of the actual rating to the estimatedrating 810. This quality metric, along with other quality metrics, may be used to fine tune future estimated ratings. - Recommendation engine 906 is further configured to periodically, and continuously if necessary, recalculate ratings and
prediction data 912. In particular, recommendation engine 906 is configured to recalculate the set of similar items for each item to reflect changes in ratings data and new ratings data. Changes in ratings data and new ratings data may be tracked and various thresholds and aging schemes used to prioritize recalculation of similar item sets. For example, items that have a relatively large amount of ratings data updated may be given priority over an item for which the calculated similar item set is simply old. It is beneficial to periodically recalculate the set of similar items for a particular item, even if no ratings data has changed for the particular item, since ratings data may have changed the set of items considered similar to the particular item. - There may be situations where limitations in the power of the computing architecture used to implement
recommendation system 904 make it difficult forrecommendation system 904 to keep up with changes in ratings data. That is, the estimated ratings maintained by recommendation engine 906 may fall behind the updated ratings data. One consequence of this situation may be a loss in accuracy of estimated ratings. There may also be situations where limitations in the available size ofnon-volatile storage 908 may make maintaining data file 914 difficult. Accordingly, according to another embodiment of the invention, the approach is segmented into a series of separate recommendation engines and separate disc arrays allowing scaling to much higher capacities and load levels.FIG. 10 depicts an example embodiment of such anarrangement 1000. -
Arrangement 1000 includes aWeb server 1002 and arecommendation system 1004, configured in accordance with an embodiment of the invention.Recommendation system 1004 includes a similars/reference engine 1006, anon-volatile storage 1008, aratings DB 1010 andrecommendation engines - A single reference users/similar items data file 1016 is used to store the ratings for
reference users 814. Reference users/similar items data file 1016 is maintained onnon-volatile storage 1008 by similars/reference engine 1006, which maintains the set ofreference users 814.Reference users 814 are chosen as described herein. If too many of users 804 qualify asreference users 814, then users 804 who have the most attractive rating patterns are chosen asreference users 814. Example criteria for choosing thereference users 814 include, without limitation, users with the most ratings, or the most ratings on otherwise unrated items, or other criteria that may be appropriate. New candidate reference users are loaded fromratings DB 1010 as they qualify, and existingreference users 814 may be displaced if any of the candidate reference users have more attractive parameters making them more suitable asreferences users 814 for making estimated ratings. Similars/reference engine 1006 is also configured to calculate sets ofsimilar items 812 for each ofitems 802 on a continuous basis and store that data into reference users/similar items data file 1016. -
Recommendation engines Recommendation engines recommendation engines Recommendation engines recommendation engines recommendation engines 1012, 1024.Recommendation engines recommendation engines recommendation engines - Reference Users' 814 active profiles of ratings and predictions are stored in one of the active user data files 1018, 1020 and estimated ratings for
reference users 814 are calculated by theappropriate recommendation engine reference user 814 updates their ratings, those ratings are also stored inratings DB 1010. Sometime later, similars/reference engine 1006 updates the shared copy of the ratings for thatreference user 814 in reference user/similar items data file 1016. As long as the rate of change of ratings forreference users 814 is small compared to the total number of ratings forreference users 814, the quality of the similar items and estimated ratings calculated using the older data will not suffer measurably.Recommendation engines - Task queues may be shared between the one or more engines, so that new work items may be triggered by one recommendation engine, but performed by a different engine if the task relates to data managed by the other engine.
- Although embodiments of the invention have been described herein in the context of using a single similars/reference engine 1006, additional scaling may be provided by using multiple similars/reference engines that are each assigned responsibility for a portion of
reference users 814. Since the determination ofsimilar items 812 requires comparing columns of ratings for pairs of items across allreference users 814, it is appropriate to splitreference users 814 by vertical segmentation by items, for example such that even/odd items (or other item segmentation) are split between different similars/reference engines. In this situation, each portion of the ratings data for eachreference user 814 is available to each similars/reference engine. More specifically, each data file is owned by a single similars/reference engine that has read/write access for maintenance, and each data file is also made available on a read-only basis to all other similars/reference engine for access to the reference ratings data. - VIII. Implementation Mechanisms
- The approach described herein for estimating how a user would rate a particular item is applicable to any type of items or content and is well suited for Internet-based rental applications for renting movies, music and games to customers. The invention may be implemented in hardware circuitry, in computer software, or a combination of hardware circuitry and computer software and is not limited to a particular hardware or software implementation. The invention is ideally suited for recommendation engine applications and may be implemented as a stand-alone mechanism, or integrated into a rental mechanism. For example, the approach is ideally suited for identifying rental items, such as movies, music and games, to recommend to users.
-
FIG. 11 is a block diagram that illustrates acomputer system 1100 upon which an embodiment of the invention may be implemented.Computer system 1100 includes abus 1102 or other communication mechanism for communicating information, and aprocessor 1104 coupled withbus 1102 for processing information.Computer system 1100 also includes amain memory 1106, such as a random access memory (RAM) or other dynamic storage device, coupled tobus 1102 for storing information and instructions to be executed byprocessor 1104.Main memory 1106 also may be used for storing temporary variables or other intermediate information during execution of instructions to be executed byprocessor 1104.Computer system 1100 further includes a read only memory (ROM) 1108 or other static storage device coupled tobus 1102 for storing static information and instructions forprocessor 1104. Astorage device 1110, such as a magnetic disk or optical disk, is provided and coupled tobus 1102 for storing information and instructions. -
Computer system 1100 may be coupled viabus 1102 to adisplay 1112, such as a cathode ray tube (CRT), for displaying information to a computer user. Aninput device 1114, including alphanumeric and other keys, is coupled tobus 1102 for communicating information and command selections toprocessor 1104. Another type of user input device iscursor control 1116, such as a mouse, a trackball, or cursor direction keys for communicating direction information and command selections toprocessor 1104 and for controlling cursor movement ondisplay 1112. This input device typically has two degrees of freedom in two axes, a first axis (e.g., x) and a second axis (e.g., y), that allows the device to specify positions in a plane. - The invention is related to the use of
computer system 1100 for implementing the techniques described herein. According to one embodiment of the invention, those techniques are performed bycomputer system 1100 in response toprocessor 1104 executing one or more sequences of one or more instructions contained inmain memory 1106. Such instructions may be read intomain memory 1106 from another machine-readable medium, such asstorage device 1110. Execution of the sequences of instructions contained inmain memory 1106 causesprocessor 1104 to perform the process steps described herein. In alternative embodiments, hard-wired circuitry may be used in place of or in combination with software instructions to implement the invention. Thus, embodiments of the invention are not limited to any specific combination of hardware circuitry and software. - The term “machine-readable medium” as used herein refers to any medium that participates in providing data that causes a machine to operation in a specific fashion. In an embodiment implemented using
computer system 1100, various machine-readable media are involved, for example, in providing instructions toprocessor 1104 for execution. Such a medium may take many forms, including but not limited to, non-volatile media, volatile media, and transmission media. Non-volatile media includes, for example, optical or magnetic disks, such asstorage device 1110. Volatile media includes dynamic memory, such asmain memory 1106. Transmission media includes coaxial cables, copper wire and fiber optics, including the wires that comprisebus 1102. Transmission media can also take the form of acoustic or light waves, such as those generated during radio wave and infrared data communications. - Common forms of machine-readable media include, for example, a floppy disk, a flexible disk, hard disk, magnetic tape, or any other magnetic medium, a CD-ROM, any other optical medium, punchcards, papertape, any other physical medium with patterns of holes, a RAM, a PROM, and EPROM, a FLASH-EPROM, any other memory chip or cartridge, a carrier wave as described hereinafter, or any other medium from which a computer can read.
- Various forms of machine-readable media may be involved in carrying one or more sequences of one or more instructions to
processor 1104 for execution. For example, the instructions may initially be carried on a magnetic disk of a remote computer. The remote computer can load the instructions into its dynamic memory and send the instructions over a telephone line using a modem. A modem local tocomputer system 1100 can receive the data on the telephone line and use an infrared transmitter to convert the data to an infrared signal. An infrared detector can receive the data carried in the infrared signal and appropriate circuitry can place the data onbus 1102.Bus 1102 carries the data tomain memory 1106, from whichprocessor 1104 retrieves and executes the instructions. The instructions received bymain memory 1106 may optionally be stored onstorage device 1110 either before or after execution byprocessor 1104. -
Computer system 1100 also includes acommunication interface 1118 coupled tobus 1102.Communication interface 1118 provides a two-way data communication coupling to anetwork link 1120 that is connected to alocal network 1122. For example,communication interface 1118 may be an integrated services digital network (ISDN) card or a modem to provide a data communication connection to a corresponding type of telephone line. As another example,communication interface 1118 may be a local area network (LAN) card to provide a data communication connection to a compatible LAN. Wireless links may also be implemented. In any such implementation,communication interface 1118 sends and receives electrical, electromagnetic or optical signals that carry digital data streams representing various types of information. -
Network link 1120 typically provides data communication through one or more networks to other data devices. For example,network link 1120 may provide a connection throughlocal network 1122 to ahost computer 1124 or to data equipment operated by an Internet Service Provider (ISP) 1126.ISP 1126 in turn provides data communication services through the worldwide packet data communication network now commonly referred to as the “Internet” 1128.Local network 1122 andInternet 1128 both use electrical, electromagnetic or optical signals that carry digital data streams. The signals through the various networks and the signals onnetwork link 1120 and throughcommunication interface 1118, which carry the digital data to and fromcomputer system 1100, are exemplary forms of carrier waves transporting the information. -
Computer system 1100 can send messages and receive data, including program code, through the network(s),network link 1120 andcommunication interface 1118. In the Internet example, aserver 1130 might transmit a requested code for an application program throughInternet 1128,ISP 1126,local network 1122 andcommunication interface 1118. The received code may be executed byprocessor 1104 as it is received, and/or stored instorage device 1110, or other non-volatile storage for later execution. In this manner,computer system 1100 may obtain application code in the form of a carrier wave. - The novel approach described herein for renting items to customers allows the decision of what items to rent to be separated from the decision of when to rent the items. Customers may specify what items to rent using the item selection criteria and receive the items at a future point in time, without having to go to the provider to pick up the items. The selection criteria may be user specific and may indicate a desired fulfillment sequence. Furthermore, customers are not constrained by conventional due dates and instead may establish continuous, serialized rental streams of items. The approach also allows more efficient inventory management.
- The “Max Out” approach for inventory management allows users to maintain their own inventory of items that are periodically replaced by other items according to specified event criteria. The event criteria that trigger sending another item to a customer are very flexible and may be tailored to the requirements of a particular application. For example, as described herein, the event criteria may include a return of any of the items currently in use by the customer or merely customer notification. This is very convenient in the context of movie rentals since the return of a movie to the provider automatically triggers the sending of another movie to the customer. The “Max Turns” approach for inventory management, when used alone or in combination with “Max Out,” provides even greater flexibility for customers and providers. The max number of turns can be selected individually for particular customers depending upon their particular needs.
- The “Max Out” and “Max Turns” approaches provide great flexibility in establishing subscription models to satisfy the needs of a particular application. Specifically, the size and replacement frequency of customer inventories can be tailored to each customer with individualized subscription plans.
- In the foregoing specification, the invention has been described as applicable to an implementation anticipating Internet based ordering and mail or other long-distance delivery of the items, where the special advantages of the method are very attractive. However the same invention may be applied in a more conventional video, games, or music rental-store setting, where subscription customers may be allowed rentals of a specified number of movies, games, or music selections at any time, and/or in one subscription period, without rental return due dates, in exchange for a periodic rental subscription fee.
- In the foregoing specification, embodiments of the invention have been described with reference to numerous specific details that may vary from implementation to implementation. Thus, the sole and exclusive indicator of what is, and is intended-by the applicants to be, the invention is the set of claims that issue from this application, in the specific form in which such claims issue, including any subsequent correction. Hence, no limitation, element, property, feature, advantage or attribute that is not expressly recited in a claim should limit the scope of such claim in any way. The specification and drawings are, accordingly, to be regarded in an illustrative rather than a restrictive sense.
Claims (109)
Priority Applications (7)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
US10/773,843 US7617127B2 (en) | 2000-04-28 | 2004-02-06 | Approach for estimating user ratings of items |
AU2005213374A AU2005213374A1 (en) | 2004-02-06 | 2005-02-04 | Approach for estimating user ratings of items |
PCT/US2005/003377 WO2005076890A2 (en) | 2004-02-06 | 2005-02-04 | Approach for estimating user ratings of items |
JP2006552232A JP2007528060A (en) | 2004-02-06 | 2005-02-04 | An approach for estimating user ratings of items |
EP05722701A EP1730686A2 (en) | 2004-02-06 | 2005-02-04 | Approach for estimating user ratings of items |
CA002556026A CA2556026A1 (en) | 2004-02-06 | 2005-02-04 | Approach for estimating user ratings of items |
US12/562,741 US8271338B2 (en) | 2004-02-06 | 2009-09-18 | Approach for estimating user ratings of items |
Applications Claiming Priority (5)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
US09/561,041 US6584450B1 (en) | 2000-04-28 | 2000-04-28 | Method and apparatus for renting items |
US21219300P | 2000-06-16 | 2000-06-16 | |
US24479300P | 2000-10-31 | 2000-10-31 | |
US09/884,816 US7403910B1 (en) | 2000-04-28 | 2001-06-18 | Approach for estimating user ratings of items |
US10/773,843 US7617127B2 (en) | 2000-04-28 | 2004-02-06 | Approach for estimating user ratings of items |
Related Parent Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
US09/884,816 Continuation-In-Part US7403910B1 (en) | 2000-04-28 | 2001-06-18 | Approach for estimating user ratings of items |
Related Child Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
US12/562,741 Continuation US8271338B2 (en) | 2004-02-06 | 2009-09-18 | Approach for estimating user ratings of items |
Publications (2)
Publication Number | Publication Date |
---|---|
US20050125307A1 true US20050125307A1 (en) | 2005-06-09 |
US7617127B2 US7617127B2 (en) | 2009-11-10 |
Family
ID=34860805
Family Applications (2)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
US10/773,843 Active 2025-06-12 US7617127B2 (en) | 2000-04-28 | 2004-02-06 | Approach for estimating user ratings of items |
US12/562,741 Active 2025-09-06 US8271338B2 (en) | 2004-02-06 | 2009-09-18 | Approach for estimating user ratings of items |
Family Applications After (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
US12/562,741 Active 2025-09-06 US8271338B2 (en) | 2004-02-06 | 2009-09-18 | Approach for estimating user ratings of items |
Country Status (6)
Country | Link |
---|---|
US (2) | US7617127B2 (en) |
EP (1) | EP1730686A2 (en) |
JP (1) | JP2007528060A (en) |
AU (1) | AU2005213374A1 (en) |
CA (1) | CA2556026A1 (en) |
WO (1) | WO2005076890A2 (en) |
Cited By (70)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20020147628A1 (en) * | 2001-02-16 | 2002-10-10 | Jeffrey Specter | Method and apparatus for generating recommendations for consumer preference items |
US20020178057A1 (en) * | 2001-05-10 | 2002-11-28 | International Business Machines Corporation | System and method for item recommendations |
US20030110056A1 (en) * | 2001-10-31 | 2003-06-12 | International Business Machines Corporation | Method for rating items within a recommendation system based on additional knowledge of item relationships |
US20030149612A1 (en) * | 2001-10-31 | 2003-08-07 | International Business Machines Corporation | Enabling a recommendation system to provide user-to-user recommendations |
US20040225577A1 (en) * | 2001-10-18 | 2004-11-11 | Gary Robinson | System and method for measuring rating reliability through rater prescience |
US20050086127A1 (en) * | 2000-04-28 | 2005-04-21 | Hastings W. R. | Approach for managing rental items across a plurality of distribution locations |
US20060136284A1 (en) * | 2004-12-17 | 2006-06-22 | Baruch Awerbuch | Recommendation system |
US20060179053A1 (en) * | 2005-02-04 | 2006-08-10 | Microsoft Corporation | Improving quality of web search results using a game |
US20060217994A1 (en) * | 2005-03-25 | 2006-09-28 | The Motley Fool, Inc. | Method and system for harnessing collective knowledge |
US20060217962A1 (en) * | 2005-03-08 | 2006-09-28 | Yasuharu Asano | Information processing device, information processing method, program, and recording medium |
US20060224442A1 (en) * | 2005-03-31 | 2006-10-05 | Round Matthew J | Closed loop voting feedback |
US20070011073A1 (en) * | 2005-03-25 | 2007-01-11 | The Motley Fool, Inc. | System, method, and computer program product for scoring items based on user sentiment and for determining the proficiency of predictors |
US20070276773A1 (en) * | 2006-03-06 | 2007-11-29 | Murali Aravamudan | Methods and systems for selecting and presenting content on a first system based on user preferences learned on a second system |
WO2008008607A2 (en) * | 2006-07-14 | 2008-01-17 | At & T Knowledge Ventures, L.P. | Method and apparatus for sharing end user feedback |
US20080071647A1 (en) * | 2003-08-13 | 2008-03-20 | Mcqueen Clyde D Iii | Personalized selection of content to present to users |
US20080071602A1 (en) * | 2006-08-31 | 2008-03-20 | Yahoo! Inc. | Enhanced user reviews |
US20080120288A1 (en) * | 2006-11-17 | 2008-05-22 | Wei Guan | Collaborative-filtering content model for recommending items |
US20080120287A1 (en) * | 2006-11-17 | 2008-05-22 | Wei Guan | Collaborative-filtering contextual model based on explicit and implicit ratings for recommending items |
US20080120339A1 (en) * | 2006-11-17 | 2008-05-22 | Wei Guan | Collaborative-filtering contextual model optimized for an objective function for recommending items |
US20080119277A1 (en) * | 2006-11-21 | 2008-05-22 | Big Fish Games, Inc. | Common Interests Affiliation Network Architecture |
US20080147500A1 (en) * | 2006-12-15 | 2008-06-19 | Malcolm Slaney | Serving advertisements using entertainment ratings in a collaborative-filtering system |
US7403910B1 (en) | 2000-04-28 | 2008-07-22 | Netflix, Inc. | Approach for estimating user ratings of items |
US20080209229A1 (en) * | 2006-11-13 | 2008-08-28 | Veveo, Inc. | Method of and system for selecting and presenting content based on user identification |
US20080243997A1 (en) * | 2007-03-29 | 2008-10-02 | Motorola, Inc. | Distributed content item recommendation system and method of operation therefor |
US7461058B1 (en) | 1999-09-24 | 2008-12-02 | Thalveg Data Flow Llc | Optimized rule based constraints for collaborative filtering systems |
US20090049082A1 (en) * | 2007-08-13 | 2009-02-19 | Yahoo! Inc. | System and method for identifying similar media objects |
US20090049091A1 (en) * | 2007-08-13 | 2009-02-19 | Yahoo! Inc. | System and method for identifying similar media objects |
US20090055376A1 (en) * | 2007-08-21 | 2009-02-26 | Yahoo! Inc. | System and method for identifying similar media objects |
US20090077132A1 (en) * | 2005-09-28 | 2009-03-19 | Sony Corporation | Information Processing Device and Method, and Program |
US20090077093A1 (en) * | 2007-09-19 | 2009-03-19 | Joydeep Sen Sarma | Feature Discretization and Cardinality Reduction Using Collaborative Filtering Techniques |
US20090076987A1 (en) * | 2004-10-18 | 2009-03-19 | Christine Chmura | System and Method for Using Artifical Intelligence in Automated Analysis of Labor Market Data |
US20090077081A1 (en) * | 2007-09-19 | 2009-03-19 | Joydeep Sen Sarma | Attribute-Based Item Similarity Using Collaborative Filtering Techniques |
US20090083126A1 (en) * | 2007-09-26 | 2009-03-26 | At&T Labs, Inc. | Methods and Apparatus for Modeling Relationships at Multiple Scales in Ratings Estimation |
US7516086B2 (en) * | 2003-09-24 | 2009-04-07 | Idearc Media Corp. | Business rating placement heuristic |
US20090132526A1 (en) * | 2007-11-19 | 2009-05-21 | Jong-Hun Park | Content recommendation apparatus and method using tag cloud |
US20090144226A1 (en) * | 2007-12-03 | 2009-06-04 | Kei Tateno | Information processing device and method, and program |
WO2009073664A2 (en) * | 2007-12-04 | 2009-06-11 | Google Inc. | Rating raters |
US20090157491A1 (en) * | 2007-12-12 | 2009-06-18 | Brougher William C | Monetization of Online Content |
US20090248494A1 (en) * | 2008-04-01 | 2009-10-01 | Certona Corporation | System and method for collecting and targeting visitor behavior |
US20090313115A1 (en) * | 2008-06-12 | 2009-12-17 | Microsoft Corporation | Sponsored search data structure |
US20090327193A1 (en) * | 2008-06-27 | 2009-12-31 | Nokia Corporation | Apparatus, method and computer program product for filtering media files |
US7788123B1 (en) | 2000-06-23 | 2010-08-31 | Ekhaus Michael A | Method and system for high performance model-based personalization |
US20100317444A1 (en) * | 2009-06-10 | 2010-12-16 | Microsoft Corporation | Using a human computation game to improve search engine performance |
US7899806B2 (en) | 2006-04-20 | 2011-03-01 | Veveo, Inc. | User interface methods and systems for selecting and presenting content based on user navigation and selection actions associated with the content |
US20110208752A1 (en) * | 2008-10-17 | 2011-08-25 | Telefonaktiebolaget Lm Ericsson (Publ) | Method and Apparatus for Default Rating Estimation |
US20110213805A1 (en) * | 2004-03-15 | 2011-09-01 | Yahoo! Inc. | Search systems and methods with integration of user annotations |
US8055553B1 (en) | 2006-01-19 | 2011-11-08 | Verizon Laboratories Inc. | Dynamic comparison text functionality |
WO2012050576A1 (en) * | 2010-10-13 | 2012-04-19 | Hewlett-Packard Development Company, L.P. | Automated negotiation |
US20120213445A1 (en) * | 2011-02-17 | 2012-08-23 | Canon Kabushiki Kaisha | Method, apparatus and system for rating images |
US20120303676A1 (en) * | 2001-02-12 | 2012-11-29 | Alexander Tuzhilin | System, Process and Software Arrangement for Providing Multidimensional Recommendations/Suggestions |
US20130041488A1 (en) * | 2007-12-14 | 2013-02-14 | John Nicholas And Kristin Gross Trust U/A/D April 13, 2010 | Integrated Gourmet Item Data Collection, Recommender and Vending System and Method |
US20130124449A1 (en) * | 2011-07-12 | 2013-05-16 | Ebay Inc. | Recommendations in a computing advice facility |
US20130173339A1 (en) * | 2011-12-30 | 2013-07-04 | Kitchology LLC | Food management services |
US20140081991A1 (en) * | 2006-12-15 | 2014-03-20 | Jeffrey Aaron | Automatic Rating Optimization |
US20140139961A1 (en) * | 2012-11-21 | 2014-05-22 | Tyrone J. Mellon, JR. | Device for providing protection against electrical shock hazards involving two circuit conductors |
US8799112B1 (en) * | 2010-12-13 | 2014-08-05 | Amazon Technologies, Inc. | Interactive map for browsing items |
US20140317106A1 (en) * | 2004-11-22 | 2014-10-23 | Facebook, Inc. | Method and apparatus for a ranking engine |
US9129008B1 (en) * | 2008-11-10 | 2015-09-08 | Google Inc. | Sentiment-based classification of media content |
US9166714B2 (en) | 2009-09-11 | 2015-10-20 | Veveo, Inc. | Method of and system for presenting enriched video viewing analytics |
US9177081B2 (en) | 2005-08-26 | 2015-11-03 | Veveo, Inc. | Method and system for processing ambiguous, multi-term search queries |
US9231897B1 (en) * | 2011-01-10 | 2016-01-05 | Amazon Technologies, Inc. | Estimated value ratings for electronic messages |
US20160044091A1 (en) * | 2013-07-31 | 2016-02-11 | Adenda Media Inc. | Extending mobile applications to the lock screen of a mobile device |
US9703779B2 (en) | 2010-02-04 | 2017-07-11 | Veveo, Inc. | Method of and system for enhanced local-device content discovery |
US10341699B2 (en) | 2004-07-30 | 2019-07-02 | Broadband Itv, Inc. | System for addressing on-demand TV program content on TV services platform of a digital TV services provider |
US10491955B2 (en) | 2004-07-30 | 2019-11-26 | Broadband Itv, Inc. | Video-on-demand content delivery system for providing video-on-demand services to TV services subscribers |
US10560733B2 (en) | 2007-06-26 | 2020-02-11 | Broadband Itv, Inc. | Dynamic adjustment of electronic program guide displays based on viewer preferences for minimizing navigation in VOD program selection |
US20200050677A1 (en) * | 2018-08-07 | 2020-02-13 | Disney Enterprises, Inc. | Joint understanding of actors, literary characters, and movies |
US10567846B2 (en) | 2007-06-26 | 2020-02-18 | Broadband Itv, Inc. | Dynamic adjustment of electronic program guide displays based on viewer preferences for minimizing navigation in VOD program selection |
US11252459B2 (en) | 2004-07-30 | 2022-02-15 | Broadband Itv, Inc. | System for addressing on-demand TV program content on TV services platform of a digital TV services provider |
US11570521B2 (en) | 2007-06-26 | 2023-01-31 | Broadband Itv, Inc. | Dynamic adjustment of electronic program guide displays based on viewer preferences for minimizing navigation in VOD program selection |
Families Citing this family (42)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
DE102005003632A1 (en) | 2005-01-20 | 2006-08-17 | Fraunhofer-Gesellschaft zur Förderung der angewandten Forschung e.V. | Catheter for the transvascular implantation of heart valve prostheses |
US8516135B2 (en) | 2005-09-30 | 2013-08-20 | Qurio Holdings, Inc. | Providing and receiving content for computer networks using a gateway and server |
US9015736B2 (en) | 2005-12-29 | 2015-04-21 | Rovi Guides, Inc. | Systems and methods for episode tracking in an interactive media environment |
US20070157222A1 (en) * | 2005-12-29 | 2007-07-05 | United Video Properties, Inc. | Systems and methods for managing content |
US20100198773A1 (en) * | 2006-11-06 | 2010-08-05 | Promethean Ventures, Llc | System and method of using movie taste for compatibility matching |
US20080228578A1 (en) * | 2007-01-25 | 2008-09-18 | Governing Dynamics, Llc | Digital rights management and data license management |
JP4247280B2 (en) * | 2007-02-14 | 2009-04-02 | 株式会社東芝 | RECOMMENDED PROGRAM INFORMATION PROVIDING DEVICE, RECOMMENDED PROGRAM INFORMATION PROVIDING METHOD, AND PROGRAM |
EP1975813A1 (en) | 2007-03-31 | 2008-10-01 | Sony Deutschland Gmbh | Method for content recommendation |
US7896915B2 (en) | 2007-04-13 | 2011-03-01 | Jenavalve Technology, Inc. | Medical device for treating a heart valve insufficiency |
US7898394B2 (en) * | 2007-05-10 | 2011-03-01 | Red Hat, Inc. | Systems and methods for community tagging |
US8266127B2 (en) * | 2007-05-31 | 2012-09-11 | Red Hat, Inc. | Systems and methods for directed forums |
US8356048B2 (en) | 2007-05-31 | 2013-01-15 | Red Hat, Inc. | Systems and methods for improved forums |
US7966319B2 (en) * | 2007-06-07 | 2011-06-21 | Red Hat, Inc. | Systems and methods for a rating system |
US8037009B2 (en) * | 2007-08-27 | 2011-10-11 | Red Hat, Inc. | Systems and methods for linking an issue with an entry in a knowledgebase |
US7840581B2 (en) * | 2008-02-01 | 2010-11-23 | Realnetworks, Inc. | Method and system for improving the quality of deep metadata associated with media content |
US9088808B1 (en) | 2008-02-08 | 2015-07-21 | Google Inc. | User interaction based related videos |
US8055655B1 (en) | 2008-02-15 | 2011-11-08 | Google Inc. | User interaction based related digital content items |
US9044318B2 (en) | 2008-02-26 | 2015-06-02 | Jenavalve Technology Gmbh | Stent for the positioning and anchoring of a valvular prosthesis |
US8317858B2 (en) | 2008-02-26 | 2012-11-27 | Jenavalve Technology, Inc. | Stent for the positioning and anchoring of a valvular prosthesis in an implantation site in the heart of a patient |
ES2903231T3 (en) | 2008-02-26 | 2022-03-31 | Jenavalve Tech Inc | Stent for positioning and anchoring a valve prosthesis at an implantation site in a patient's heart |
WO2010122448A1 (en) * | 2009-04-20 | 2010-10-28 | Koninklijke Philips Electronics N.V. | Method and system for rating items |
JP2013526388A (en) | 2010-05-25 | 2013-06-24 | イエナバルブ テクノロジー インク | Artificial heart valve, and transcatheter delivery prosthesis comprising an artificial heart valve and a stent |
US8533066B2 (en) | 2010-10-13 | 2013-09-10 | Hulu, LLC | Method and apparatus for recommending media programs based on correlated user feedback |
US9256886B2 (en) * | 2010-10-25 | 2016-02-09 | Microsoft Technology Licensing, Llc | Content recommendation system and method |
US9990651B2 (en) | 2010-11-17 | 2018-06-05 | Amobee, Inc. | Method and apparatus for selective delivery of ads based on factors including site clustering |
US20120303415A1 (en) * | 2011-05-25 | 2012-11-29 | Ari Edelson | System and method of providing recommendations |
US10475050B2 (en) | 2011-07-08 | 2019-11-12 | Netflix, Inc. | Identifying similar items based on global interaction history |
US10311386B2 (en) * | 2011-07-08 | 2019-06-04 | Netflix, Inc. | Identifying similar items based on interaction history |
US20130173491A1 (en) * | 2011-12-28 | 2013-07-04 | Target Brands, Inc. | Highlighting guest reviews |
US8751429B2 (en) | 2012-07-09 | 2014-06-10 | Wine Ring, Inc. | Personal taste assessment method and system |
US8843951B1 (en) * | 2012-08-27 | 2014-09-23 | Google Inc. | User behavior indicator |
US20140172499A1 (en) * | 2012-12-17 | 2014-06-19 | United Video Properties, Inc. | Systems and methods providing content ratings based on environmental factors |
AU2013378061A1 (en) | 2013-02-14 | 2015-08-20 | Ringit, Inc. | Recommendation system based on group profiles of personal taste |
US9305303B2 (en) | 2013-05-07 | 2016-04-05 | Nasdaq, Inc. | Webcast systems and methods with audience sentiment feedback and analysis |
JP6563394B2 (en) | 2013-08-30 | 2019-08-21 | イェーナヴァルヴ テクノロジー インコーポレイテッド | Radially foldable frame for an artificial valve and method for manufacturing the frame |
US9264656B2 (en) | 2014-02-26 | 2016-02-16 | Rovi Guides, Inc. | Systems and methods for managing storage space |
US20150302425A1 (en) * | 2014-04-22 | 2015-10-22 | International Business Machines Corporation | Assigning priority levels to citizen sensor reports |
US9400780B2 (en) | 2014-10-17 | 2016-07-26 | International Business Machines Corporation | Perspective data management for common features of multiple items |
US10709555B2 (en) | 2015-05-01 | 2020-07-14 | Jenavalve Technology, Inc. | Device and method with reduced pacemaker rate in heart valve replacement |
CN105760443B (en) | 2016-02-03 | 2017-11-21 | 广州市动景计算机科技有限公司 | Item recommendation system, project recommendation device and item recommendation method |
EP3454795B1 (en) | 2016-05-13 | 2023-01-11 | JenaValve Technology, Inc. | Heart valve prosthesis delivery system for delivery of heart valve prosthesis with introducer sheath and loading system |
CN108664484A (en) * | 2017-03-28 | 2018-10-16 | 腾讯科技(北京)有限公司 | Media content recommendations method and device |
Citations (40)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US4839505A (en) * | 1986-05-29 | 1989-06-13 | Videomat Associates | Apparatus and method for storing and retrieving articles |
US5095195A (en) * | 1988-08-03 | 1992-03-10 | Thru-The-Wall Corporation | Automated videocassette dispensing terminal with reservation feature |
US5664110A (en) * | 1994-12-08 | 1997-09-02 | Highpoint Systems, Inc. | Remote ordering system |
US5704017A (en) * | 1996-02-16 | 1997-12-30 | Microsoft Corporation | Collaborative filtering utilizing a belief network |
US5790426A (en) * | 1996-04-30 | 1998-08-04 | Athenium L.L.C. | Automated collaborative filtering system |
US5812641A (en) * | 1994-10-28 | 1998-09-22 | Nippon T.M.I. Co., Ltd. | Method of renting portable-type communicating devices |
US5842199A (en) * | 1996-10-18 | 1998-11-24 | Regents Of The University Of Minnesota | System, method and article of manufacture for using receiver operating curves to evaluate predictive utility |
US5867799A (en) * | 1996-04-04 | 1999-02-02 | Lang; Andrew K. | Information system and method for filtering a massive flow of information entities to meet user information classification needs |
US5909638A (en) * | 1996-08-06 | 1999-06-01 | Maximum Video Systems, Inc. | High speed video distribution and manufacturing system |
US5918213A (en) * | 1995-12-22 | 1999-06-29 | Mci Communications Corporation | System and method for automated remote previewing and purchasing of music, video, software, and other multimedia products |
US5918014A (en) * | 1995-12-27 | 1999-06-29 | Athenium, L.L.C. | Automated collaborative filtering in world wide web advertising |
US6012052A (en) * | 1998-01-15 | 2000-01-04 | Microsoft Corporation | Methods and apparatus for building resource transition probability models for use in pre-fetching resources, editing resource link topology, building resource link topology templates, and collaborative filtering |
US6016475A (en) * | 1996-10-08 | 2000-01-18 | The Regents Of The University Of Minnesota | System, method, and article of manufacture for generating implicit ratings based on receiver operating curves |
US6041311A (en) * | 1995-06-30 | 2000-03-21 | Microsoft Corporation | Method and apparatus for item recommendation using automated collaborative filtering |
US6049777A (en) * | 1995-06-30 | 2000-04-11 | Microsoft Corporation | Computer-implemented collaborative filtering based method for recommending an item to a user |
US6092049A (en) * | 1995-06-30 | 2000-07-18 | Microsoft Corporation | Method and apparatus for efficiently recommending items using automated collaborative filtering and feature-guided automated collaborative filtering |
US6108493A (en) * | 1996-10-08 | 2000-08-22 | Regents Of The University Of Minnesota | System, method, and article of manufacture for utilizing implicit ratings in collaborative filters |
US6112186A (en) * | 1995-06-30 | 2000-08-29 | Microsoft Corporation | Distributed system for facilitating exchange of user information and opinion using automated collaborative filtering |
US6266649B1 (en) * | 1998-09-18 | 2001-07-24 | Amazon.Com, Inc. | Collaborative recommendations using item-to-item similarity mappings |
US6330592B1 (en) * | 1998-12-05 | 2001-12-11 | Vignette Corporation | Method, memory, product, and code for displaying pre-customized content associated with visitor data |
US20040153413A1 (en) * | 2003-01-31 | 2004-08-05 | Gross John N. | Notification system and method for media Queue |
US20040158504A1 (en) * | 2003-01-31 | 2004-08-12 | Gross John N. | Method of providing access to playable media |
US20040158503A1 (en) * | 2003-01-31 | 2004-08-12 | Gross John N. | Media queue monitor |
US20040162783A1 (en) * | 2003-01-31 | 2004-08-19 | Gross John N. | Media queue replenisher |
US6782370B1 (en) * | 1997-09-04 | 2004-08-24 | Cendant Publishing, Inc. | System and method for providing recommendation of goods or services based on recorded purchasing history |
US20040172275A1 (en) * | 2003-01-31 | 2004-09-02 | Gross John N. | Media exchange system & method |
US20040172342A1 (en) * | 2003-01-31 | 2004-09-02 | Gross John N. | System for providing access to playable media |
US20040172274A1 (en) * | 2003-01-31 | 2004-09-02 | Gross John N. | Media auto exchange system and method |
US6799165B1 (en) * | 1998-07-28 | 2004-09-28 | Eimar M. Boesjes | Apparatus and methods for inventory, sale, and delivery of digitally transferable goods |
US20040243480A1 (en) * | 2003-05-28 | 2004-12-02 | Gross John N. | Method of controlling electronic commerce queue |
US20040243479A1 (en) * | 2003-05-28 | 2004-12-02 | Gross John N. | Method of monitoring electronic commerce queue |
US20040267669A1 (en) * | 2003-06-25 | 2004-12-30 | Gross John N. | Method of processing rental requests and returns |
US20050096127A1 (en) * | 2001-08-23 | 2005-05-05 | Dabrowski Stanley P. | Method and apparatus for autonomous validation of issued scrip media |
US6996606B2 (en) * | 2001-10-05 | 2006-02-07 | Nihon Digital Co., Ltd. | Junk mail rejection system |
US7012520B2 (en) * | 2003-06-17 | 2006-03-14 | Infraegis, Inc. | Global intelligent remote detection system |
US7054834B2 (en) * | 2000-04-11 | 2006-05-30 | Nec Corporation | Online distribution system and method |
US7058581B1 (en) * | 1999-12-20 | 2006-06-06 | Ward Kraft, Inc. | System and method of distributing and returning products |
US7055741B2 (en) * | 2002-10-31 | 2006-06-06 | United Parcel Service Of America, Inc. | Systems and methods of inventory management utilizing unattended facilities |
US7174568B2 (en) * | 2001-01-31 | 2007-02-06 | Sony Computer Entertainment America Inc. | Method and system for securely distributing computer software products |
US7181758B1 (en) * | 1994-07-25 | 2007-02-20 | Data Innovation, L.L.C. | Information distribution and processing system |
Family Cites Families (26)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US4967906A (en) | 1987-10-19 | 1990-11-06 | Diebold, Incorporated | Apparatus for dispensing and accepting return of reusable articles |
WO1991005307A1 (en) | 1989-09-29 | 1991-04-18 | Image Engineering, Inc. | A method of evaluating consumer choice through concept testing for the marketing and development of consumer products |
JP2642315B2 (en) | 1994-08-22 | 1997-08-20 | 日機装株式会社 | Blood purification device |
US5872850A (en) * | 1996-02-02 | 1999-02-16 | Microsoft Corporation | System for enabling information marketplace |
JP3116851B2 (en) * | 1997-02-24 | 2000-12-11 | 日本電気株式会社 | Information filtering method and apparatus |
US6807537B1 (en) | 1997-12-04 | 2004-10-19 | Microsoft Corporation | Mixtures of Bayesian networks |
JPH11282875A (en) * | 1998-03-27 | 1999-10-15 | Oki Electric Ind Co Ltd | Information filtering device |
US6321221B1 (en) | 1998-07-17 | 2001-11-20 | Net Perceptions, Inc. | System, method and article of manufacture for increasing the user value of recommendations |
US6334127B1 (en) | 1998-07-17 | 2001-12-25 | Net Perceptions, Inc. | System, method and article of manufacture for making serendipity-weighted recommendations to a user |
US6286139B1 (en) | 1998-08-04 | 2001-09-04 | Teluve Corporation | Internet-based video ordering system and method |
US6236985B1 (en) | 1998-10-07 | 2001-05-22 | International Business Machines Corporation | System and method for searching databases with applications such as peer groups, collaborative filtering, and e-commerce |
US6412012B1 (en) | 1998-12-23 | 2002-06-25 | Net Perceptions, Inc. | System, method, and article of manufacture for making a compatibility-aware recommendations to a user |
US6487541B1 (en) | 1999-01-22 | 2002-11-26 | International Business Machines Corporation | System and method for collaborative filtering with applications to e-commerce |
US6308168B1 (en) | 1999-02-09 | 2001-10-23 | Knowledge Discovery One, Inc. | Metadata-driven data presentation module for database system |
US6321179B1 (en) | 1999-06-29 | 2001-11-20 | Xerox Corporation | System and method for using noisy collaborative filtering to rank and present items |
US6389372B1 (en) | 1999-06-29 | 2002-05-14 | Xerox Corporation | System and method for bootstrapping a collaborative filtering system |
US6487539B1 (en) | 1999-08-06 | 2002-11-26 | International Business Machines Corporation | Semantic based collaborative filtering |
JP2001243285A (en) | 2000-02-28 | 2001-09-07 | Oki Electric Ind Co Ltd | Method for lending rental merchandise, its collecting method and merchandise rental system |
US6751736B1 (en) | 2000-03-14 | 2004-06-15 | International Business Machines Corporation | Method and apparatus for E-commerce by using optional fields for virtual bar codes |
US7546252B2 (en) | 2000-04-28 | 2009-06-09 | Netflix, Inc. | Approach for managing rental items across a plurality of distribution locations |
US6655963B1 (en) * | 2000-07-31 | 2003-12-02 | Microsoft Corporation | Methods and apparatus for predicting and selectively collecting preferences based on personality diagnosis |
JP2002074142A (en) * | 2000-08-30 | 2002-03-15 | Toyota Motor Corp | Merchandise recommendation system |
US6484123B2 (en) | 2000-11-30 | 2002-11-19 | International Business Machines Corporation | Method and system to identify which predictors are important for making a forecast with a collaborative filter |
JP2003167901A (en) * | 2001-11-29 | 2003-06-13 | Kddi Corp | Method of cooperative filtering, device for cooperative filtering and program for cooperative filtering |
JP2005122369A (en) * | 2003-10-15 | 2005-05-12 | Kddi Corp | Item recommendation degree presenting apparatus for recommender system |
US7590546B2 (en) | 2003-10-22 | 2009-09-15 | Chuang Thomas C | System and method for renting or purchasing goods via a communications network |
-
2004
- 2004-02-06 US US10/773,843 patent/US7617127B2/en active Active
-
2005
- 2005-02-04 EP EP05722701A patent/EP1730686A2/en not_active Withdrawn
- 2005-02-04 CA CA002556026A patent/CA2556026A1/en not_active Abandoned
- 2005-02-04 AU AU2005213374A patent/AU2005213374A1/en not_active Abandoned
- 2005-02-04 JP JP2006552232A patent/JP2007528060A/en active Pending
- 2005-02-04 WO PCT/US2005/003377 patent/WO2005076890A2/en active Application Filing
-
2009
- 2009-09-18 US US12/562,741 patent/US8271338B2/en active Active
Patent Citations (45)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US4839505A (en) * | 1986-05-29 | 1989-06-13 | Videomat Associates | Apparatus and method for storing and retrieving articles |
US5095195A (en) * | 1988-08-03 | 1992-03-10 | Thru-The-Wall Corporation | Automated videocassette dispensing terminal with reservation feature |
US7181758B1 (en) * | 1994-07-25 | 2007-02-20 | Data Innovation, L.L.C. | Information distribution and processing system |
US5812641A (en) * | 1994-10-28 | 1998-09-22 | Nippon T.M.I. Co., Ltd. | Method of renting portable-type communicating devices |
US5664110A (en) * | 1994-12-08 | 1997-09-02 | Highpoint Systems, Inc. | Remote ordering system |
US6041311A (en) * | 1995-06-30 | 2000-03-21 | Microsoft Corporation | Method and apparatus for item recommendation using automated collaborative filtering |
US6112186A (en) * | 1995-06-30 | 2000-08-29 | Microsoft Corporation | Distributed system for facilitating exchange of user information and opinion using automated collaborative filtering |
US6092049A (en) * | 1995-06-30 | 2000-07-18 | Microsoft Corporation | Method and apparatus for efficiently recommending items using automated collaborative filtering and feature-guided automated collaborative filtering |
US6049777A (en) * | 1995-06-30 | 2000-04-11 | Microsoft Corporation | Computer-implemented collaborative filtering based method for recommending an item to a user |
US5918213A (en) * | 1995-12-22 | 1999-06-29 | Mci Communications Corporation | System and method for automated remote previewing and purchasing of music, video, software, and other multimedia products |
US5918014A (en) * | 1995-12-27 | 1999-06-29 | Athenium, L.L.C. | Automated collaborative filtering in world wide web advertising |
US5704017A (en) * | 1996-02-16 | 1997-12-30 | Microsoft Corporation | Collaborative filtering utilizing a belief network |
US5983214A (en) * | 1996-04-04 | 1999-11-09 | Lycos, Inc. | System and method employing individual user content-based data and user collaborative feedback data to evaluate the content of an information entity in a large information communication network |
US6029161A (en) * | 1996-04-04 | 2000-02-22 | Lycos, Inc. | Multi-level mindpool system especially adapted to provide collaborative filter data for a large scale information filtering system |
US5867799A (en) * | 1996-04-04 | 1999-02-02 | Lang; Andrew K. | Information system and method for filtering a massive flow of information entities to meet user information classification needs |
US5884282A (en) * | 1996-04-30 | 1999-03-16 | Robinson; Gary B. | Automated collaborative filtering system |
US5790426A (en) * | 1996-04-30 | 1998-08-04 | Athenium L.L.C. | Automated collaborative filtering system |
US5909638A (en) * | 1996-08-06 | 1999-06-01 | Maximum Video Systems, Inc. | High speed video distribution and manufacturing system |
US6016475A (en) * | 1996-10-08 | 2000-01-18 | The Regents Of The University Of Minnesota | System, method, and article of manufacture for generating implicit ratings based on receiver operating curves |
US6108493A (en) * | 1996-10-08 | 2000-08-22 | Regents Of The University Of Minnesota | System, method, and article of manufacture for utilizing implicit ratings in collaborative filters |
US5842199A (en) * | 1996-10-18 | 1998-11-24 | Regents Of The University Of Minnesota | System, method and article of manufacture for using receiver operating curves to evaluate predictive utility |
US6782370B1 (en) * | 1997-09-04 | 2004-08-24 | Cendant Publishing, Inc. | System and method for providing recommendation of goods or services based on recorded purchasing history |
US6012052A (en) * | 1998-01-15 | 2000-01-04 | Microsoft Corporation | Methods and apparatus for building resource transition probability models for use in pre-fetching resources, editing resource link topology, building resource link topology templates, and collaborative filtering |
US6799165B1 (en) * | 1998-07-28 | 2004-09-28 | Eimar M. Boesjes | Apparatus and methods for inventory, sale, and delivery of digitally transferable goods |
US6266649B1 (en) * | 1998-09-18 | 2001-07-24 | Amazon.Com, Inc. | Collaborative recommendations using item-to-item similarity mappings |
US6330592B1 (en) * | 1998-12-05 | 2001-12-11 | Vignette Corporation | Method, memory, product, and code for displaying pre-customized content associated with visitor data |
US7058581B1 (en) * | 1999-12-20 | 2006-06-06 | Ward Kraft, Inc. | System and method of distributing and returning products |
US7054834B2 (en) * | 2000-04-11 | 2006-05-30 | Nec Corporation | Online distribution system and method |
US7174568B2 (en) * | 2001-01-31 | 2007-02-06 | Sony Computer Entertainment America Inc. | Method and system for securely distributing computer software products |
US20050096127A1 (en) * | 2001-08-23 | 2005-05-05 | Dabrowski Stanley P. | Method and apparatus for autonomous validation of issued scrip media |
US6996606B2 (en) * | 2001-10-05 | 2006-02-07 | Nihon Digital Co., Ltd. | Junk mail rejection system |
US7055741B2 (en) * | 2002-10-31 | 2006-06-06 | United Parcel Service Of America, Inc. | Systems and methods of inventory management utilizing unattended facilities |
US20040172342A1 (en) * | 2003-01-31 | 2004-09-02 | Gross John N. | System for providing access to playable media |
US20040153413A1 (en) * | 2003-01-31 | 2004-08-05 | Gross John N. | Notification system and method for media Queue |
US20040158504A1 (en) * | 2003-01-31 | 2004-08-12 | Gross John N. | Method of providing access to playable media |
US20040172274A1 (en) * | 2003-01-31 | 2004-09-02 | Gross John N. | Media auto exchange system and method |
US20040162783A1 (en) * | 2003-01-31 | 2004-08-19 | Gross John N. | Media queue replenisher |
US20040172275A1 (en) * | 2003-01-31 | 2004-09-02 | Gross John N. | Media exchange system & method |
US20040158503A1 (en) * | 2003-01-31 | 2004-08-12 | Gross John N. | Media queue monitor |
US20040243479A1 (en) * | 2003-05-28 | 2004-12-02 | Gross John N. | Method of monitoring electronic commerce queue |
US20040243480A1 (en) * | 2003-05-28 | 2004-12-02 | Gross John N. | Method of controlling electronic commerce queue |
US7046138B2 (en) * | 2003-06-17 | 2006-05-16 | Intelagents, Inc. | System and method for monitoring a security of an asset |
US7012520B2 (en) * | 2003-06-17 | 2006-03-14 | Infraegis, Inc. | Global intelligent remote detection system |
US20060149685A1 (en) * | 2003-06-25 | 2006-07-06 | Gross John N | Method of processing rental requests and returns |
US20040267669A1 (en) * | 2003-06-25 | 2004-12-30 | Gross John N. | Method of processing rental requests and returns |
Cited By (194)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US8548987B2 (en) | 1999-09-24 | 2013-10-01 | Thalveg Data Flow Llc | System and method for efficiently providing a recommendation |
US7461058B1 (en) | 1999-09-24 | 2008-12-02 | Thalveg Data Flow Llc | Optimized rule based constraints for collaborative filtering systems |
US20050086127A1 (en) * | 2000-04-28 | 2005-04-21 | Hastings W. R. | Approach for managing rental items across a plurality of distribution locations |
US7403910B1 (en) | 2000-04-28 | 2008-07-22 | Netflix, Inc. | Approach for estimating user ratings of items |
US7546252B2 (en) | 2000-04-28 | 2009-06-09 | Netflix, Inc. | Approach for managing rental items across a plurality of distribution locations |
US7788123B1 (en) | 2000-06-23 | 2010-08-31 | Ekhaus Michael A | Method and system for high performance model-based personalization |
US8155992B2 (en) | 2000-06-23 | 2012-04-10 | Thalveg Data Flow Llc | Method and system for high performance model-based personalization |
US20100332408A1 (en) * | 2000-06-23 | 2010-12-30 | Ekhaus Michael A | Method and System for High Performance Model-Based Personalization |
US8984000B2 (en) * | 2001-02-12 | 2015-03-17 | Facebook, Inc. | System, process and software arrangement for providing multidimensional recommendations/suggestions |
US20120303569A1 (en) * | 2001-02-12 | 2012-11-29 | Alexander Tuzhilin | System, Process and Software Arrangement for Providing Multidimensional Recommendations/Suggestions |
US20120303676A1 (en) * | 2001-02-12 | 2012-11-29 | Alexander Tuzhilin | System, Process and Software Arrangement for Providing Multidimensional Recommendations/Suggestions |
US8924264B2 (en) * | 2001-02-12 | 2014-12-30 | Facebook, Inc. | System, process and software arrangement for providing multidimensional recommendations/suggestions |
US8930243B2 (en) | 2001-02-12 | 2015-01-06 | Facebook, Inc. | System, process and software arrangement for providing multidimensional recommendations/suggestions |
US20080065469A1 (en) * | 2001-02-16 | 2008-03-13 | Specter Jeffrey P | Method and apparatus for generating recommendations for consumer preference items |
US20020147628A1 (en) * | 2001-02-16 | 2002-10-10 | Jeffrey Specter | Method and apparatus for generating recommendations for consumer preference items |
US11157984B2 (en) * | 2001-05-10 | 2021-10-26 | International Business Machines Corporation | System and method for item recommendations |
US8700448B2 (en) * | 2001-05-10 | 2014-04-15 | International Business Machines Corporation | System and method for item recommendations |
US20140081996A1 (en) * | 2001-05-10 | 2014-03-20 | International Business Machines Corporation | System and method for item recommendations |
US20020178057A1 (en) * | 2001-05-10 | 2002-11-28 | International Business Machines Corporation | System and method for item recommendations |
US20040225577A1 (en) * | 2001-10-18 | 2004-11-11 | Gary Robinson | System and method for measuring rating reliability through rater prescience |
US20030110056A1 (en) * | 2001-10-31 | 2003-06-12 | International Business Machines Corporation | Method for rating items within a recommendation system based on additional knowledge of item relationships |
US20030149612A1 (en) * | 2001-10-31 | 2003-08-07 | International Business Machines Corporation | Enabling a recommendation system to provide user-to-user recommendations |
US20080071647A1 (en) * | 2003-08-13 | 2008-03-20 | Mcqueen Clyde D Iii | Personalized selection of content to present to users |
US7536320B2 (en) * | 2003-08-13 | 2009-05-19 | Amazon.Com, Inc. | Method, system, and computer readable medium for the selection of content items for presentation to users |
US20080133298A1 (en) * | 2003-08-13 | 2008-06-05 | Mcqueen Clyde D | Personalization using multiple personalized selection algorithms |
US7765130B2 (en) | 2003-08-13 | 2010-07-27 | Amazon.Com, Inc. | Personalization using multiple personalized selection algorithms |
US8224704B2 (en) | 2003-09-24 | 2012-07-17 | SuperMedia LLC | Business rating placement heuristic |
US7516086B2 (en) * | 2003-09-24 | 2009-04-07 | Idearc Media Corp. | Business rating placement heuristic |
US9489463B2 (en) * | 2004-03-15 | 2016-11-08 | Excalibur Ip, Llc | Search systems and methods with integration of user annotations |
US9984164B2 (en) | 2004-03-15 | 2018-05-29 | Excalibur Ip, Llc | Search systems and methods with integration of aggregate user annotations |
US8930246B2 (en) | 2004-03-15 | 2015-01-06 | Verizon Patent And Licensing Inc. | Dynamic comparison text functionality |
US20110213805A1 (en) * | 2004-03-15 | 2011-09-01 | Yahoo! Inc. | Search systems and methods with integration of user annotations |
US10555014B2 (en) | 2004-07-30 | 2020-02-04 | Broadband Itv, Inc. | System for addressing on-demand TV program content on TV services platform of a digital TV services provider |
US10506269B2 (en) | 2004-07-30 | 2019-12-10 | Broadband Itv, Inc. | System for addressing on-demand TV program content on TV services platform of a digital TV services provider |
US11259089B2 (en) | 2004-07-30 | 2022-02-22 | Broadband Itv, Inc. | Video-on-demand content delivery method for providing video-on-demand services to TV service subscribers |
US10791351B2 (en) | 2004-07-30 | 2020-09-29 | Broadband Itv, Inc. | System for addressing on-demand TV program content on TV services platform of a digital TV services provider |
US10785517B2 (en) | 2004-07-30 | 2020-09-22 | Broadband Itv, Inc. | Method for addressing on-demand TV program content on TV services platform of a digital TV services provider |
US11272233B2 (en) | 2004-07-30 | 2022-03-08 | Broadband Itv, Inc. | System for addressing on-demand TV program content on TV services platform of a digital TV services provider |
US11516525B2 (en) | 2004-07-30 | 2022-11-29 | Broadband Itv, Inc. | System for addressing on-demand TV program content on TV services platform of a digital TV services provider |
US11259060B2 (en) | 2004-07-30 | 2022-02-22 | Broadband Itv, Inc. | System for addressing on-demand TV program content on TV services platform of a digital TV services provider |
US10536750B2 (en) | 2004-07-30 | 2020-01-14 | Broadband Itv, Inc. | Video-on-demand content delivery system for providing video-on-demand services to TV service subscribers |
US11601697B2 (en) | 2004-07-30 | 2023-03-07 | Broadband Itv, Inc. | System for addressing on-demand TV program content on TV services platform of a digital TV services provider |
US10536751B2 (en) | 2004-07-30 | 2020-01-14 | Broadband Itv, Inc. | Video-on-demand content delivery system for providing video-on-demand services to TV service subscribers |
US10893334B2 (en) | 2004-07-30 | 2021-01-12 | Broadband Itv, Inc. | Video-on-demand content delivery method for providing video-on-demand services to TV service subscribers |
US10491954B2 (en) | 2004-07-30 | 2019-11-26 | Broadband Itv, Inc. | Video-on-demand content delivery method for providing video-on-demand services to TV service subscribers |
US10491955B2 (en) | 2004-07-30 | 2019-11-26 | Broadband Itv, Inc. | Video-on-demand content delivery system for providing video-on-demand services to TV services subscribers |
US10349101B2 (en) | 2004-07-30 | 2019-07-09 | Broadband Itv, Inc. | System for addressing on-demand TV program content on TV services platform of a digital TV services provider |
US10341699B2 (en) | 2004-07-30 | 2019-07-02 | Broadband Itv, Inc. | System for addressing on-demand TV program content on TV services platform of a digital TV services provider |
US11259059B2 (en) | 2004-07-30 | 2022-02-22 | Broadband Itv, Inc. | System for addressing on-demand TV program content on TV services platform of a digital TV services provider |
US11252459B2 (en) | 2004-07-30 | 2022-02-15 | Broadband Itv, Inc. | System for addressing on-demand TV program content on TV services platform of a digital TV services provider |
US11252476B2 (en) | 2004-07-30 | 2022-02-15 | Broadband Itv, Inc. | Video-on-demand content delivery system for providing video-on-demand services to TV service subscribers |
US20090076987A1 (en) * | 2004-10-18 | 2009-03-19 | Christine Chmura | System and Method for Using Artifical Intelligence in Automated Analysis of Labor Market Data |
US20140317106A1 (en) * | 2004-11-22 | 2014-10-23 | Facebook, Inc. | Method and apparatus for a ranking engine |
US20060136284A1 (en) * | 2004-12-17 | 2006-06-22 | Baruch Awerbuch | Recommendation system |
US7603343B2 (en) * | 2005-02-04 | 2009-10-13 | Microsoft Corporation | Quality of web search results using a game |
US20060179053A1 (en) * | 2005-02-04 | 2006-08-10 | Microsoft Corporation | Improving quality of web search results using a game |
US7720849B2 (en) * | 2005-03-08 | 2010-05-18 | Sony Corporation | Information processing device, information processing method, program, and recording medium |
US20060217962A1 (en) * | 2005-03-08 | 2006-09-28 | Yasuharu Asano | Information processing device, information processing method, program, and recording medium |
US20060218179A1 (en) * | 2005-03-25 | 2006-09-28 | The Motley Fool, Inc. | System, method, and computer program product for scoring items based on user sentiment and for determining the proficiency of predictors |
US7813986B2 (en) | 2005-03-25 | 2010-10-12 | The Motley Fool, Llc | System, method, and computer program product for scoring items based on user sentiment and for determining the proficiency of predictors |
US20060217994A1 (en) * | 2005-03-25 | 2006-09-28 | The Motley Fool, Inc. | Method and system for harnessing collective knowledge |
US20070011073A1 (en) * | 2005-03-25 | 2007-01-11 | The Motley Fool, Inc. | System, method, and computer program product for scoring items based on user sentiment and for determining the proficiency of predictors |
US7882006B2 (en) | 2005-03-25 | 2011-02-01 | The Motley Fool, Llc | System, method, and computer program product for scoring items based on user sentiment and for determining the proficiency of predictors |
US8566144B2 (en) * | 2005-03-31 | 2013-10-22 | Amazon Technologies, Inc. | Closed loop voting feedback |
US20060224442A1 (en) * | 2005-03-31 | 2006-10-05 | Round Matthew J | Closed loop voting feedback |
US9177081B2 (en) | 2005-08-26 | 2015-11-03 | Veveo, Inc. | Method and system for processing ambiguous, multi-term search queries |
US20090077132A1 (en) * | 2005-09-28 | 2009-03-19 | Sony Corporation | Information Processing Device and Method, and Program |
US8117211B2 (en) * | 2005-09-28 | 2012-02-14 | Sony Corporation | Information processing device and method, and program |
US8055553B1 (en) | 2006-01-19 | 2011-11-08 | Verizon Laboratories Inc. | Dynamic comparison text functionality |
US20110154264A1 (en) * | 2006-03-06 | 2011-06-23 | Veveo, Inc. | Methods and Systems for Selecting and Presenting Content Based on Learned Periodicity of User Content Selection |
US20070276773A1 (en) * | 2006-03-06 | 2007-11-29 | Murali Aravamudan | Methods and systems for selecting and presenting content on a first system based on user preferences learned on a second system |
US8073848B2 (en) | 2006-03-06 | 2011-12-06 | Veveo, Inc. | Methods and systems for selecting and presenting content based on user preference information extracted from an aggregate preference signature |
US9213755B2 (en) | 2006-03-06 | 2015-12-15 | Veveo, Inc. | Methods and systems for selecting and presenting content based on context sensitive user preferences |
US8543516B2 (en) | 2006-03-06 | 2013-09-24 | Veveo, Inc. | Methods and systems for selecting and presenting content on a first system based on user preferences learned on a second system |
US8112454B2 (en) | 2006-03-06 | 2012-02-07 | Veveo, Inc. | Methods and systems for ordering content items according to learned user preferences |
US9092503B2 (en) | 2006-03-06 | 2015-07-28 | Veveo, Inc. | Methods and systems for selecting and presenting content based on dynamically identifying microgenres associated with the content |
US8583566B2 (en) * | 2006-03-06 | 2013-11-12 | Veveo, Inc. | Methods and systems for selecting and presenting content based on learned periodicity of user content selection |
US9075861B2 (en) | 2006-03-06 | 2015-07-07 | Veveo, Inc. | Methods and systems for segmenting relative user preferences into fine-grain and coarse-grain collections |
US8156113B2 (en) | 2006-03-06 | 2012-04-10 | Veveo, Inc. | Methods and systems for selecting and presenting content based on dynamically identifying microgenres associated with the content |
US7949627B2 (en) * | 2006-03-06 | 2011-05-24 | Veveo, Inc. | Methods and systems for selecting and presenting content based on learned periodicity of user content selection |
US8478794B2 (en) | 2006-03-06 | 2013-07-02 | Veveo, Inc. | Methods and systems for segmenting relative user preferences into fine-grain and coarse-grain collections |
US8949231B2 (en) | 2006-03-06 | 2015-02-03 | Veveo, Inc. | Methods and systems for selecting and presenting content based on activity level spikes associated with the content |
US7835998B2 (en) | 2006-03-06 | 2010-11-16 | Veveo, Inc. | Methods and systems for selecting and presenting content on a first system based on user preferences learned on a second system |
US8943083B2 (en) | 2006-03-06 | 2015-01-27 | Veveo, Inc. | Methods and systems for segmenting relative user preferences into fine-grain and coarse-grain collections |
US8438160B2 (en) | 2006-03-06 | 2013-05-07 | Veveo, Inc. | Methods and systems for selecting and presenting content based on dynamically identifying Microgenres Associated with the content |
US8429155B2 (en) | 2006-03-06 | 2013-04-23 | Veveo, Inc. | Methods and systems for selecting and presenting content based on activity level spikes associated with the content |
US9128987B2 (en) | 2006-03-06 | 2015-09-08 | Veveo, Inc. | Methods and systems for selecting and presenting content based on a comparison of preference signatures from multiple users |
US8429188B2 (en) | 2006-03-06 | 2013-04-23 | Veveo, Inc. | Methods and systems for selecting and presenting content based on context sensitive user preferences |
US7885904B2 (en) | 2006-03-06 | 2011-02-08 | Veveo, Inc. | Methods and systems for selecting and presenting content on a first system based on user preferences learned on a second system |
US8825576B2 (en) | 2006-03-06 | 2014-09-02 | Veveo, Inc. | Methods and systems for selecting and presenting content on a first system based on user preferences learned on a second system |
US8380726B2 (en) | 2006-03-06 | 2013-02-19 | Veveo, Inc. | Methods and systems for selecting and presenting content based on a comparison of preference signatures from multiple users |
US8688746B2 (en) | 2006-04-20 | 2014-04-01 | Veveo, Inc. | User interface methods and systems for selecting and presenting content based on user relationships |
US8375069B2 (en) | 2006-04-20 | 2013-02-12 | Veveo Inc. | User interface methods and systems for selecting and presenting content based on user navigation and selection actions associated with the content |
US8423583B2 (en) | 2006-04-20 | 2013-04-16 | Veveo Inc. | User interface methods and systems for selecting and presenting content based on user relationships |
US7899806B2 (en) | 2006-04-20 | 2011-03-01 | Veveo, Inc. | User interface methods and systems for selecting and presenting content based on user navigation and selection actions associated with the content |
US10146840B2 (en) | 2006-04-20 | 2018-12-04 | Veveo, Inc. | User interface methods and systems for selecting and presenting content based on user relationships |
US8086602B2 (en) | 2006-04-20 | 2011-12-27 | Veveo Inc. | User interface methods and systems for selecting and presenting content based on user navigation and selection actions associated with the content |
US9087109B2 (en) | 2006-04-20 | 2015-07-21 | Veveo, Inc. | User interface methods and systems for selecting and presenting content based on user relationships |
WO2008008607A3 (en) * | 2006-07-14 | 2009-02-05 | At & T Knowledge Ventures Lp | Method and apparatus for sharing end user feedback |
WO2008008607A2 (en) * | 2006-07-14 | 2008-01-17 | At & T Knowledge Ventures, L.P. | Method and apparatus for sharing end user feedback |
US7885863B2 (en) * | 2006-08-31 | 2011-02-08 | Yahoo! Inc. | Methods and computer readable media for enhancing product reviews |
US20080071602A1 (en) * | 2006-08-31 | 2008-03-20 | Yahoo! Inc. | Enhanced user reviews |
US20080209229A1 (en) * | 2006-11-13 | 2008-08-28 | Veveo, Inc. | Method of and system for selecting and presenting content based on user identification |
US8078884B2 (en) | 2006-11-13 | 2011-12-13 | Veveo, Inc. | Method of and system for selecting and presenting content based on user identification |
US7590616B2 (en) * | 2006-11-17 | 2009-09-15 | Yahoo! Inc. | Collaborative-filtering contextual model based on explicit and implicit ratings for recommending items |
US20080120287A1 (en) * | 2006-11-17 | 2008-05-22 | Wei Guan | Collaborative-filtering contextual model based on explicit and implicit ratings for recommending items |
US7584171B2 (en) * | 2006-11-17 | 2009-09-01 | Yahoo! Inc. | Collaborative-filtering content model for recommending items |
US20080120339A1 (en) * | 2006-11-17 | 2008-05-22 | Wei Guan | Collaborative-filtering contextual model optimized for an objective function for recommending items |
US20080120288A1 (en) * | 2006-11-17 | 2008-05-22 | Wei Guan | Collaborative-filtering content model for recommending items |
US7574422B2 (en) * | 2006-11-17 | 2009-08-11 | Yahoo! Inc. | Collaborative-filtering contextual model optimized for an objective function for recommending items |
US20080119277A1 (en) * | 2006-11-21 | 2008-05-22 | Big Fish Games, Inc. | Common Interests Affiliation Network Architecture |
US20080147500A1 (en) * | 2006-12-15 | 2008-06-19 | Malcolm Slaney | Serving advertisements using entertainment ratings in a collaborative-filtering system |
US20140081991A1 (en) * | 2006-12-15 | 2014-03-20 | Jeffrey Aaron | Automatic Rating Optimization |
US10028000B2 (en) | 2006-12-15 | 2018-07-17 | At&T Intellectual Property I, L.P. | Automatic rating optimization |
US9456250B2 (en) * | 2006-12-15 | 2016-09-27 | At&T Intellectual Property I, L.P. | Automatic rating optimization |
US11589093B2 (en) | 2007-03-12 | 2023-02-21 | Broadband Itv, Inc. | System for addressing on-demand TV program content on TV services platform of a digital TV services provider |
US11245942B2 (en) | 2007-03-12 | 2022-02-08 | Broadband Itv, Inc. | Method for addressing on-demand TV program content on TV services platform of a digital TV services provider |
US20080243997A1 (en) * | 2007-03-29 | 2008-10-02 | Motorola, Inc. | Distributed content item recommendation system and method of operation therefor |
US11570500B2 (en) | 2007-06-26 | 2023-01-31 | Broadband Itv, Inc. | Dynamic adjustment of electronic program guide displays based on viewer preferences for minimizing navigation in VOD program selection |
US10560733B2 (en) | 2007-06-26 | 2020-02-11 | Broadband Itv, Inc. | Dynamic adjustment of electronic program guide displays based on viewer preferences for minimizing navigation in VOD program selection |
US11265589B2 (en) | 2007-06-26 | 2022-03-01 | Broadband Itv, Inc. | Dynamic adjustment of electronic program guide displays based on viewer preferences for minimizing navigation in VOD program selection |
US10623793B2 (en) | 2007-06-26 | 2020-04-14 | Broadband Itv, Inc. | Dynamic adjustment of electronic program guide displays based on viewer preferences for minimizing navigation in VOD program selection |
US11570521B2 (en) | 2007-06-26 | 2023-01-31 | Broadband Itv, Inc. | Dynamic adjustment of electronic program guide displays based on viewer preferences for minimizing navigation in VOD program selection |
US11695976B2 (en) | 2007-06-26 | 2023-07-04 | Broadband Itv, Inc. | Dynamic adjustment of electronic program guide displays based on viewer preferences for minimizing navigation in VOD program selection |
US11272235B2 (en) | 2007-06-26 | 2022-03-08 | Broadband Itv, Inc. | Dynamic adjustment of electronic program guide displays based on viewer preferences for minimizing navigation in VOD program selection |
US11290763B2 (en) | 2007-06-26 | 2022-03-29 | Broadband Itv, Inc. | Dynamic adjustment of electronic program guide displays based on viewer preferences for minimizing navigation in VOD program selection |
US10582243B2 (en) | 2007-06-26 | 2020-03-03 | Broadband Itv, Inc. | Dynamic adjustment of electronic program guide displays based on viewer preferences for minimizing navigation in VOD program selection |
US11582498B2 (en) | 2007-06-26 | 2023-02-14 | Broadband Itv, Inc. | Dynamic adjustment of electronic program guide displays based on viewer preferences for minimizing navigation in VOD program selection |
US10567846B2 (en) | 2007-06-26 | 2020-02-18 | Broadband Itv, Inc. | Dynamic adjustment of electronic program guide displays based on viewer preferences for minimizing navigation in VOD program selection |
US11277669B2 (en) | 2007-06-26 | 2022-03-15 | Broadband Itv, Inc. | Dynamic adjustment of electronic program guide displays based on viewer preferences for minimizing navigation in VOD program selection |
US20090049082A1 (en) * | 2007-08-13 | 2009-02-19 | Yahoo! Inc. | System and method for identifying similar media objects |
US20090049091A1 (en) * | 2007-08-13 | 2009-02-19 | Yahoo! Inc. | System and method for identifying similar media objects |
US7849092B2 (en) * | 2007-08-13 | 2010-12-07 | Yahoo! Inc. | System and method for identifying similar media objects |
US8407230B2 (en) * | 2007-08-13 | 2013-03-26 | Yahoo! Inc. | System and method for identifying similar media objects |
US10152517B2 (en) | 2007-08-13 | 2018-12-11 | Excalibur Ip, Llc | System and method for identifying similar media objects |
US20090055376A1 (en) * | 2007-08-21 | 2009-02-26 | Yahoo! Inc. | System and method for identifying similar media objects |
US20090077093A1 (en) * | 2007-09-19 | 2009-03-19 | Joydeep Sen Sarma | Feature Discretization and Cardinality Reduction Using Collaborative Filtering Techniques |
US20090077081A1 (en) * | 2007-09-19 | 2009-03-19 | Joydeep Sen Sarma | Attribute-Based Item Similarity Using Collaborative Filtering Techniques |
US20090083126A1 (en) * | 2007-09-26 | 2009-03-26 | At&T Labs, Inc. | Methods and Apparatus for Modeling Relationships at Multiple Scales in Ratings Estimation |
US8463662B2 (en) | 2007-09-26 | 2013-06-11 | At&T Intellectual Property I, L.P. | Methods and apparatus for modeling relationships at multiple scales in ratings estimation |
US8229798B2 (en) * | 2007-09-26 | 2012-07-24 | At&T Intellectual Property I, L.P. | Methods and apparatus for modeling relationships at multiple scales in ratings estimation |
US8209337B2 (en) * | 2007-11-19 | 2012-06-26 | Core Logic, Inc. | Content recommendation apparatus and method using tag cloud |
US20090132526A1 (en) * | 2007-11-19 | 2009-05-21 | Jong-Hun Park | Content recommendation apparatus and method using tag cloud |
US20090144226A1 (en) * | 2007-12-03 | 2009-06-04 | Kei Tateno | Information processing device and method, and program |
WO2009073664A3 (en) * | 2007-12-04 | 2009-08-13 | Google Inc | Rating raters |
WO2009073664A2 (en) * | 2007-12-04 | 2009-06-11 | Google Inc. | Rating raters |
US8150842B2 (en) | 2007-12-12 | 2012-04-03 | Google Inc. | Reputation of an author of online content |
US9760547B1 (en) | 2007-12-12 | 2017-09-12 | Google Inc. | Monetization of online content |
US20090157490A1 (en) * | 2007-12-12 | 2009-06-18 | Justin Lawyer | Credibility of an Author of Online Content |
US20090157491A1 (en) * | 2007-12-12 | 2009-06-18 | Brougher William C | Monetization of Online Content |
US8645396B2 (en) | 2007-12-12 | 2014-02-04 | Google Inc. | Reputation scoring of an author |
US8291492B2 (en) | 2007-12-12 | 2012-10-16 | Google Inc. | Authentication of a contributor of online content |
US20090165128A1 (en) * | 2007-12-12 | 2009-06-25 | Mcnally Michael David | Authentication of a Contributor of Online Content |
US8126882B2 (en) | 2007-12-12 | 2012-02-28 | Google Inc. | Credibility of an author of online content |
US8626608B2 (en) * | 2007-12-14 | 2014-01-07 | John Nicholas and Kristin Gross Trust | Recommendation systems using gourmet item sampling events |
US10482484B2 (en) | 2007-12-14 | 2019-11-19 | John Nicholas And Kristin Gross Trust U/A/D April 13, 2010 | Item data collection systems and methods with social network integration |
US9037515B2 (en) | 2007-12-14 | 2015-05-19 | John Nicholas and Kristin Gross | Social networking websites and systems for publishing sampling event data |
US20130041488A1 (en) * | 2007-12-14 | 2013-02-14 | John Nicholas And Kristin Gross Trust U/A/D April 13, 2010 | Integrated Gourmet Item Data Collection, Recommender and Vending System and Method |
US20090248494A1 (en) * | 2008-04-01 | 2009-10-01 | Certona Corporation | System and method for collecting and targeting visitor behavior |
US10878480B2 (en) | 2008-04-01 | 2020-12-29 | Certona Corporation | System and method for collecting and targeting visitor behavior |
US9916611B2 (en) * | 2008-04-01 | 2018-03-13 | Certona Corporation | System and method for collecting and targeting visitor behavior |
US8606627B2 (en) * | 2008-06-12 | 2013-12-10 | Microsoft Corporation | Sponsored search data structure |
US20090313115A1 (en) * | 2008-06-12 | 2009-12-17 | Microsoft Corporation | Sponsored search data structure |
US20090327193A1 (en) * | 2008-06-27 | 2009-12-31 | Nokia Corporation | Apparatus, method and computer program product for filtering media files |
US20110208752A1 (en) * | 2008-10-17 | 2011-08-25 | Telefonaktiebolaget Lm Ericsson (Publ) | Method and Apparatus for Default Rating Estimation |
US8429175B2 (en) * | 2008-10-17 | 2013-04-23 | Telefonaktiebolaget Lm Ericsson (Publ) | Method and apparatus for default rating estimation |
US10191905B2 (en) * | 2008-11-10 | 2019-01-29 | Google Llc | Sentiment-based classification of media content |
US10698942B2 (en) | 2008-11-10 | 2020-06-30 | Google Llc | Sentiment-based classification of media content |
US10956482B2 (en) | 2008-11-10 | 2021-03-23 | Google Llc | Sentiment-based classification of media content |
US9129008B1 (en) * | 2008-11-10 | 2015-09-08 | Google Inc. | Sentiment-based classification of media content |
US9495425B1 (en) | 2008-11-10 | 2016-11-15 | Google Inc. | Sentiment-based classification of media content |
US11379512B2 (en) | 2008-11-10 | 2022-07-05 | Google Llc | Sentiment-based classification of media content |
US9875244B1 (en) | 2008-11-10 | 2018-01-23 | Google Llc | Sentiment-based classification of media content |
US20100317444A1 (en) * | 2009-06-10 | 2010-12-16 | Microsoft Corporation | Using a human computation game to improve search engine performance |
US8285706B2 (en) * | 2009-06-10 | 2012-10-09 | Microsoft Corporation | Using a human computation game to improve search engine performance |
US9166714B2 (en) | 2009-09-11 | 2015-10-20 | Veveo, Inc. | Method of and system for presenting enriched video viewing analytics |
US9703779B2 (en) | 2010-02-04 | 2017-07-11 | Veveo, Inc. | Method of and system for enhanced local-device content discovery |
WO2012050576A1 (en) * | 2010-10-13 | 2012-04-19 | Hewlett-Packard Development Company, L.P. | Automated negotiation |
US8799112B1 (en) * | 2010-12-13 | 2014-08-05 | Amazon Technologies, Inc. | Interactive map for browsing items |
US9231897B1 (en) * | 2011-01-10 | 2016-01-05 | Amazon Technologies, Inc. | Estimated value ratings for electronic messages |
US20120213445A1 (en) * | 2011-02-17 | 2012-08-23 | Canon Kabushiki Kaisha | Method, apparatus and system for rating images |
US9009096B2 (en) * | 2011-07-12 | 2015-04-14 | Ebay Inc. | Recommendations in a computing advice facility |
US20150186383A1 (en) * | 2011-07-12 | 2015-07-02 | Ebay Inc. | Recommendations in a computing advice facility |
US10318534B2 (en) * | 2011-07-12 | 2019-06-11 | Paypal, Inc. | Recommendations in a computing advice facility |
US20130124449A1 (en) * | 2011-07-12 | 2013-05-16 | Ebay Inc. | Recommendations in a computing advice facility |
US20130173339A1 (en) * | 2011-12-30 | 2013-07-04 | Kitchology LLC | Food management services |
US20140139961A1 (en) * | 2012-11-21 | 2014-05-22 | Tyrone J. Mellon, JR. | Device for providing protection against electrical shock hazards involving two circuit conductors |
US20160044091A1 (en) * | 2013-07-31 | 2016-02-11 | Adenda Media Inc. | Extending mobile applications to the lock screen of a mobile device |
US9923953B2 (en) * | 2013-07-31 | 2018-03-20 | Adenda Media Inc. | Extending mobile applications to the lock screen of a mobile device |
US10834174B2 (en) | 2013-07-31 | 2020-11-10 | Adenda Media Inc. | Systems and methods for displaying ads on a mobile device outside of a mobile app |
US11212340B2 (en) | 2013-07-31 | 2021-12-28 | Adenda Media Inc. | Extending mobile applications to the lock screen of a mobile device |
US11652869B2 (en) | 2013-07-31 | 2023-05-16 | Adenda Media Inc. | Systems and methods for displaying ads on a mobile device outside of a mobile app |
US11785074B2 (en) | 2013-07-31 | 2023-10-10 | Adenda Media Inc. | Extending mobile applications to the lock screen of a mobile device |
US20200050677A1 (en) * | 2018-08-07 | 2020-02-13 | Disney Enterprises, Inc. | Joint understanding of actors, literary characters, and movies |
Also Published As
Publication number | Publication date |
---|---|
EP1730686A2 (en) | 2006-12-13 |
US20100010877A1 (en) | 2010-01-14 |
CA2556026A1 (en) | 2005-08-25 |
JP2007528060A (en) | 2007-10-04 |
WO2005076890A3 (en) | 2008-11-13 |
US8271338B2 (en) | 2012-09-18 |
US7617127B2 (en) | 2009-11-10 |
AU2005213374A1 (en) | 2005-08-25 |
WO2005076890A2 (en) | 2005-08-25 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
US7617127B2 (en) | Approach for estimating user ratings of items | |
US7546252B2 (en) | Approach for managing rental items across a plurality of distribution locations | |
US7024381B1 (en) | Approach for renting items to customers | |
US8972290B2 (en) | Rental inventory management | |
Montgomery et al. | Designing a better shopbot | |
JP3389948B2 (en) | Display ad selection system | |
US10922646B1 (en) | Multi-echelon inventory planning under dynamic fulfillment policies | |
US20050197855A1 (en) | Method and apparatus for renting items | |
US20140258042A1 (en) | System and Method for Facilitating Real Estate Transactions | |
US11934987B2 (en) | Sentient optimization for continuous supply chain management | |
US20080015943A1 (en) | Order fulfillment logic for a field service system | |
US20230108410A1 (en) | Method and apparatus for managing allocations of media content in electronic segments | |
CN102132577A (en) | Model-based resource allocation | |
US20140337304A1 (en) | Application retention metrics | |
JP2016201141A (en) | Replying of estimated value of retrieval keyword for all accounts | |
US20210185000A1 (en) | Dynamic Email Content Engine | |
Shi | Optimal matchmaking strategy in two-sided marketplaces | |
Xie et al. | The benefits of delay to online decision-making | |
US20050283376A1 (en) | Service level agreements supporting apparatus | |
KR20070015918A (en) | Approach for estimating user ratings of items | |
Çavdar et al. | Capacity allocation in a service system with preferred service completion times | |
Manshadi et al. | Redesigning VolunteerMatch's Ranking Algorithm: Toward More Equitable Access to Volunteers | |
Tang et al. | Online learning and matching for multiproduct systems with general upgrading | |
Jiang | Online advertisements and multi-armed bandits | |
JP4334198B2 (en) | Production planning apparatus, production planning method, recording medium thereof, and program |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
AS | Assignment |
Owner name: NETFLIX, INC., CALIFORNIA Free format text: ASSIGNMENT OF ASSIGNORS INTEREST;ASSIGNORS:HUNT, NEIL DUNCAN;LANNING, STANLEY MIGUEL;HASTINGS, W. REED;AND OTHERS;REEL/FRAME:015502/0986 Effective date: 20040617 |
|
STCF | Information on status: patent grant |
Free format text: PATENTED CASE |
|
FPAY | Fee payment |
Year of fee payment: 4 |
|
FPAY | Fee payment |
Year of fee payment: 8 |
|
MAFP | Maintenance fee payment |
Free format text: PAYMENT OF MAINTENANCE FEE, 12TH YEAR, LARGE ENTITY (ORIGINAL EVENT CODE: M1553); ENTITY STATUS OF PATENT OWNER: LARGE ENTITY Year of fee payment: 12 |